client_golang/prometheus/http.go

506 lines
13 KiB
Go
Raw Normal View History

// Copyright 2014 The Prometheus Authors
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package prometheus
import (
"bufio"
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
"compress/gzip"
"io"
"net"
"net/http"
"strconv"
"strings"
"sync"
"time"
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
"github.com/prometheus/common/expfmt"
)
// TODO(beorn7): Remove this whole file. It is a partial mirror of
// promhttp/http.go (to avoid circular import chains) where everything HTTP
// related should live. The functions here are just for avoiding
// breakage. Everything is deprecated.
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
const (
contentTypeHeader = "Content-Type"
contentEncodingHeader = "Content-Encoding"
acceptEncodingHeader = "Accept-Encoding"
)
var gzipPool = sync.Pool{
New: func() interface{} {
return gzip.NewWriter(nil)
},
}
2016-08-04 16:26:27 +03:00
// Handler returns an HTTP handler for the DefaultGatherer. It is
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
// already instrumented with InstrumentHandler (using "prometheus" as handler
// name).
//
// Deprecated: Please note the issues described in the doc comment of
// InstrumentHandler. You might want to consider using promhttp.Handler instead.
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
func Handler() http.Handler {
return InstrumentHandler("prometheus", UninstrumentedHandler())
}
2016-08-04 16:26:27 +03:00
// UninstrumentedHandler returns an HTTP handler for the DefaultGatherer.
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
//
// Deprecated: Use promhttp.HandlerFor(DefaultGatherer, promhttp.HandlerOpts{})
// instead. See there for further documentation.
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
func UninstrumentedHandler() http.Handler {
return http.HandlerFunc(func(rsp http.ResponseWriter, req *http.Request) {
2016-08-04 16:26:27 +03:00
mfs, err := DefaultGatherer.Gather()
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
if err != nil {
httpError(rsp, err)
return
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
}
contentType := expfmt.Negotiate(req.Header)
header := rsp.Header()
header.Set(contentTypeHeader, string(contentType))
w := io.Writer(rsp)
if gzipAccepted(req.Header) {
header.Set(contentEncodingHeader, "gzip")
gz := gzipPool.Get().(*gzip.Writer)
defer gzipPool.Put(gz)
gz.Reset(w)
defer gz.Close()
w = gz
}
enc := expfmt.NewEncoder(w, contentType)
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
for _, mf := range mfs {
if err := enc.Encode(mf); err != nil {
httpError(rsp, err)
return
Create a public registry interface and separate out HTTP exposition General context and approch =========================== This is the first part of the long awaited wider refurbishment of `client_golang/prometheus/...`. After a lot of struggling, I decided to not go for one breaking big-bang, but cut things into smaller steps after all, mostly to keep the changes manageable and easy to review. I'm aiming for having the invasive breaking changes concentrated in as few steps as possible (ideally one). Some steps will not be breaking at all, but typically there will be breaking changes that only affect quite special cases so that 95+% of users will not be affected. This first step is an example for that, see details below. What's happening in this commit? ================================ This step is about finally creating an exported registry interface. This could not be done by simply export the existing internal implementation because the interface would be _way_ too fat. This commit introduces a qutie lean `Registry` interface (compared to the previous interval implementation). The functions that act on the default registry are retained (with very few exceptions) so that most use cases won't see a change. However, several of those are deprecated now to clean up the namespace in the future. The default registry is kept in the public variable `DefaultRegistry`. This follows the example of the http package in the standard library (cf. `http.DefaultServeMux`, `http.DefaultClient`) with the same implications. (This pattern is somewhat disputed within the Go community but I chose to go with the devil you know instead of creating something more complex or even disallowing any changes to the default registry. The current approach gives everybody the freedom to not touch DefaultRegistry or to do everything with a custom registry to play save.) Another important part in making the registry lean is the extraction of the HTTP exposition, which also allows for customization of the HTTP exposition. Note that the separation of metric collection and exposition has the side effect that managing the MetricFamily and Metric protobuf objects in a free-list or pool isn't really feasible anymore. By now (with better GC in more recent Go versions), the returns were anyway dimisishing. To be effective at all, scrapes had to happen more often than GC cycles, and even then most elements of the protobufs (everything excetp the MetricFamily and Metric structs themselves) would still cause allocation churn. In a future breaking change, the signature of the Write method in the Metric interface will be adjusted accordingly. In this commit, avoiding breakage is more important. The following issues are fixed by this commit (some solved "on the fly" now that I was touching the code anyway and it would have been stupid to port the bugs): https://github.com/prometheus/client_golang/issues/46 https://github.com/prometheus/client_golang/issues/100 https://github.com/prometheus/client_golang/issues/170 https://github.com/prometheus/client_golang/issues/205 Documentation including examples have been amended as required. What future changes does this commit enable? ============================================ The following items are not yet implemented, but this commit opens the possibility of implementing these independently. - The separation of the HTTP exposition allows the implementation of other exposition methods based on the Registry interface, as known from other Prometheus client libraries, e.g. sending the metrics to Graphite. Cf. https://github.com/prometheus/client_golang/issues/197 - The public `Registry` interface allows the implementation of convenience tools for testing metrics collection. Those tools can inspect the collected MetricFamily protobufs and compare them to expectation. Also, tests can use their own testing instance of a registry. Cf. https://github.com/prometheus/client_golang/issues/58 Notable non-goals of this commit ================================ Non-goals that will be tackled later ------------------------------------ The following two issues are quite closely connected to the changes in this commit but the line has been drawn deliberately to address them in later steps of the refurbishment: - `InstrumentHandler` has many known problems. The plan is to create a saner way to conveniently intrument HTTP handlers and remove the old `InstrumentHandler` altogether. To keep breakage low for now, even the default handler to expose metrics is still using the old `InstrumentHandler`. This leads to weird naming inconsistencies but I have deemed it better to not break the world right now but do it in the change that provides better ways of instrumenting HTTP handlers. Cf. https://github.com/prometheus/client_golang/issues/200 - There is work underway to make the whole handling of metric descriptors (`Desc`) more intuitive and transparent for the user (including an ability for less strict checking, cf. https://github.com/prometheus/client_golang/issues/47). That's quite invasive from the perspective of the internal code, namely the registry. I deliberately kept those changes out of this commit. - While this commit adds new external dependency, the effort to vendor anything within the library that is not visible in any exported types will have to be done later. Non-goals that _might_ be tackled later --------------------------------------- There is a strong and understandable urge to divide the `prometheus` package into a number of sub-packages (like `registry`, `collectors`, `http`, `metrics`, …). However, to not run into a multitude of circular import chains, this would need to break every single existing usage of the library. (As just one example, if the ubiquitious `prometheus.MustRegister` (with more than 2,000 uses on GitHub alone) is kept in the `prometheus` package, but the other registry concerns go into a new `registry` package, then the `prometheus` package would import the `registry` package (to call the actual register method), while at the same time the `registry` package needs to import the `prometheus` package to access `Collector`, `Metric`, `Desc` and more. If we moved `MustRegister` into the `registry` package, thousands of code lines would have to be fixed (which would be easy if the world was a mono repo, but it is not). If we moved everything else the proposed registry package needs into packages of their own, we would break thousands of other code lines.) The main problem is really the top-level functions like `MustRegister`, `Handler`, …, which effectively pull everything into one package. Those functions are however very convenient for the easy and very frequent use-cases. This problem has to be revisited later. For now, I'm trying to keep the amount of exported names in the package as low as possible (e.g. I unexported expvarCollector in this commit because the NewExpvarCollector constructor is enough to export, and it is now consistent with other collectors, like the goCollector). Non-goals that won't be tackled anytime soon -------------------------------------------- Something that I have played with a lot is "streaming collection", i.e. allow an implementation of the `Registry` interface that collects metrics incrementally and serves them while doing so. As it has turned out, this has many many issues and makes the `Registry` interface very clunky. Eventually, I made the call that it is unlikely we will really implement streaming collection; and making the interface more clunky for something that might not even happen is really a big no-no. Note that the `Registry` interface only creates the in-memory representation of the metric family protobufs in one go. The serializaton onto the wire can still be handled in a streaming fashion (which hasn't been done so far, without causing any trouble, but might be done in the future without breaking any interfaces). What are the breaking changes? ============================== - Signatures of functions pushing to Pushgateway have changed to allow arbitrary grouping (which was planned for a long time anyway, and now that I had to work on the Push code anyway for the registry refurbishment, I finally did it, cf. https://github.com/prometheus/client_golang/issues/100). With the gained insight that pushing to the default registry is almost never the right thing, and now that we are breaking the Push call anyway, all the Push functions were moved to their own package, which cleans up the namespace and is more idiomatic (pushing Collectors is now literally done by `push.Collectors(...)`). - The registry is doing more consistency checks by default now. Past creators of inconsistent metrics could have masked the problem by not setting `EnableCollectChecks`. Those inconsistencies will now be detected. (But note that a "best effort" metrics collection is now possible with `HandlerOpts.ErrorHandling = ContinueOnError`.) - `EnableCollectChecks` is gone. The registry is now performing some of those checks anyway (see previous item), and a registry with all of those checks can now be created with `NewPedanticRegistry` (only used for testing). - `PanicOnCollectError` is gone. This behavior can now be configured when creating a custom HTTP handler.
2016-07-20 18:11:14 +03:00
}
}
})
}
var instLabels = []string{"method", "code"}
type nower interface {
Now() time.Time
}
type nowFunc func() time.Time
func (n nowFunc) Now() time.Time {
return n()
}
var now nower = nowFunc(func() time.Time {
return time.Now()
})
// InstrumentHandler wraps the given HTTP handler for instrumentation. It
// registers four metric collectors (if not already done) and reports HTTP
2015-01-13 19:26:38 +03:00
// metrics to the (newly or already) registered collectors: http_requests_total
// (CounterVec), http_request_duration_microseconds (Summary),
// http_request_size_bytes (Summary), http_response_size_bytes (Summary). Each
// has a constant label named "handler" with the provided handlerName as
// value. http_requests_total is a metric vector partitioned by HTTP method
// (label name "method") and HTTP status code (label name "code").
//
// Deprecated: InstrumentHandler has several issues. Use the tooling provided in
// package promhttp instead. The issues are the following: (1) It uses Summaries
// rather than Histograms. Summaries are not useful if aggregation across
// multiple instances is required. (2) It uses microseconds as unit, which is
// deprecated and should be replaced by seconds. (3) The size of the request is
// calculated in a separate goroutine. Since this calculator requires access to
// the request header, it creates a race with any writes to the header performed
// during request handling. httputil.ReverseProxy is a prominent example for a
// handler performing such writes. (4) It has additional issues with HTTP/2, cf.
// https://github.com/prometheus/client_golang/issues/272.
func InstrumentHandler(handlerName string, handler http.Handler) http.HandlerFunc {
return InstrumentHandlerFunc(handlerName, handler.ServeHTTP)
}
// InstrumentHandlerFunc wraps the given function for instrumentation. It
// otherwise works in the same way as InstrumentHandler (and shares the same
// issues).
//
// Deprecated: InstrumentHandlerFunc is deprecated for the same reasons as
// InstrumentHandler is. Use the tooling provided in package promhttp instead.
func InstrumentHandlerFunc(handlerName string, handlerFunc func(http.ResponseWriter, *http.Request)) http.HandlerFunc {
return InstrumentHandlerFuncWithOpts(
SummaryOpts{
Subsystem: "http",
ConstLabels: Labels{"handler": handlerName},
Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
},
handlerFunc,
)
}
// InstrumentHandlerWithOpts works like InstrumentHandler (and shares the same
// issues) but provides more flexibility (at the cost of a more complex call
// syntax). As InstrumentHandler, this function registers four metric
// collectors, but it uses the provided SummaryOpts to create them. However, the
// fields "Name" and "Help" in the SummaryOpts are ignored. "Name" is replaced
// by "requests_total", "request_duration_microseconds", "request_size_bytes",
// and "response_size_bytes", respectively. "Help" is replaced by an appropriate
2015-01-13 19:26:38 +03:00
// help string. The names of the variable labels of the http_requests_total
// CounterVec are "method" (get, post, etc.), and "code" (HTTP status code).
//
// If InstrumentHandlerWithOpts is called as follows, it mimics exactly the
// behavior of InstrumentHandler:
//
// prometheus.InstrumentHandlerWithOpts(
// prometheus.SummaryOpts{
// Subsystem: "http",
// ConstLabels: prometheus.Labels{"handler": handlerName},
// },
// handler,
// )
//
// Technical detail: "requests_total" is a CounterVec, not a SummaryVec, so it
// cannot use SummaryOpts. Instead, a CounterOpts struct is created internally,
// and all its fields are set to the equally named fields in the provided
// SummaryOpts.
//
// Deprecated: InstrumentHandlerWithOpts is deprecated for the same reasons as
// InstrumentHandler is. Use the tooling provided in package promhttp instead.
func InstrumentHandlerWithOpts(opts SummaryOpts, handler http.Handler) http.HandlerFunc {
return InstrumentHandlerFuncWithOpts(opts, handler.ServeHTTP)
}
// InstrumentHandlerFuncWithOpts works like InstrumentHandlerFunc (and shares
// the same issues) but provides more flexibility (at the cost of a more complex
// call syntax). See InstrumentHandlerWithOpts for details how the provided
// SummaryOpts are used.
//
// Deprecated: InstrumentHandlerFuncWithOpts is deprecated for the same reasons
// as InstrumentHandler is. Use the tooling provided in package promhttp instead.
func InstrumentHandlerFuncWithOpts(opts SummaryOpts, handlerFunc func(http.ResponseWriter, *http.Request)) http.HandlerFunc {
reqCnt := NewCounterVec(
CounterOpts{
Namespace: opts.Namespace,
Subsystem: opts.Subsystem,
Name: "requests_total",
Help: "Total number of HTTP requests made.",
ConstLabels: opts.ConstLabels,
},
instLabels,
)
if err := Register(reqCnt); err != nil {
if are, ok := err.(AlreadyRegisteredError); ok {
reqCnt = are.ExistingCollector.(*CounterVec)
} else {
panic(err)
}
}
opts.Name = "request_duration_microseconds"
opts.Help = "The HTTP request latencies in microseconds."
reqDur := NewSummary(opts)
if err := Register(reqDur); err != nil {
if are, ok := err.(AlreadyRegisteredError); ok {
reqDur = are.ExistingCollector.(Summary)
} else {
panic(err)
}
}
opts.Name = "request_size_bytes"
opts.Help = "The HTTP request sizes in bytes."
reqSz := NewSummary(opts)
if err := Register(reqSz); err != nil {
if are, ok := err.(AlreadyRegisteredError); ok {
reqSz = are.ExistingCollector.(Summary)
} else {
panic(err)
}
}
opts.Name = "response_size_bytes"
opts.Help = "The HTTP response sizes in bytes."
resSz := NewSummary(opts)
if err := Register(resSz); err != nil {
if are, ok := err.(AlreadyRegisteredError); ok {
resSz = are.ExistingCollector.(Summary)
} else {
panic(err)
}
}
return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
now := time.Now()
delegate := &responseWriterDelegator{ResponseWriter: w}
2016-09-10 18:24:13 +03:00
out := computeApproximateRequestSize(r)
_, cn := w.(http.CloseNotifier)
_, fl := w.(http.Flusher)
_, hj := w.(http.Hijacker)
_, rf := w.(io.ReaderFrom)
var rw http.ResponseWriter
if cn && fl && hj && rf {
rw = &fancyResponseWriterDelegator{delegate}
} else {
rw = delegate
}
handlerFunc(rw, r)
elapsed := float64(time.Since(now)) / float64(time.Microsecond)
method := sanitizeMethod(r.Method)
code := sanitizeCode(delegate.status)
reqCnt.WithLabelValues(method, code).Inc()
reqDur.Observe(elapsed)
resSz.Observe(float64(delegate.written))
reqSz.Observe(float64(<-out))
})
}
2016-09-10 18:24:13 +03:00
func computeApproximateRequestSize(r *http.Request) <-chan int {
// Get URL length in current goroutine for avoiding a race condition.
2016-09-10 18:24:13 +03:00
// HandlerFunc that runs in parallel may modify the URL.
s := 0
if r.URL != nil {
s += len(r.URL.String())
}
2016-09-10 18:24:13 +03:00
out := make(chan int, 1)
2016-09-10 18:24:13 +03:00
go func() {
s += len(r.Method)
s += len(r.Proto)
for name, values := range r.Header {
s += len(name)
for _, value := range values {
s += len(value)
}
}
s += len(r.Host)
// N.B. r.Form and r.MultipartForm are assumed to be included in r.URL.
if r.ContentLength != -1 {
s += int(r.ContentLength)
}
out <- s
close(out)
}()
return out
}
type responseWriterDelegator struct {
http.ResponseWriter
status int
written int64
wroteHeader bool
}
func (r *responseWriterDelegator) WriteHeader(code int) {
r.status = code
r.wroteHeader = true
r.ResponseWriter.WriteHeader(code)
}
func (r *responseWriterDelegator) Write(b []byte) (int, error) {
if !r.wroteHeader {
r.WriteHeader(http.StatusOK)
}
n, err := r.ResponseWriter.Write(b)
r.written += int64(n)
return n, err
}
type fancyResponseWriterDelegator struct {
*responseWriterDelegator
}
func (f *fancyResponseWriterDelegator) CloseNotify() <-chan bool {
//lint:ignore SA1019 http.CloseNotifier is deprecated but we don't want to
//remove support from client_golang yet.
return f.ResponseWriter.(http.CloseNotifier).CloseNotify()
}
func (f *fancyResponseWriterDelegator) Flush() {
f.ResponseWriter.(http.Flusher).Flush()
}
func (f *fancyResponseWriterDelegator) Hijack() (net.Conn, *bufio.ReadWriter, error) {
return f.ResponseWriter.(http.Hijacker).Hijack()
}
func (f *fancyResponseWriterDelegator) ReadFrom(r io.Reader) (int64, error) {
if !f.wroteHeader {
f.WriteHeader(http.StatusOK)
}
n, err := f.ResponseWriter.(io.ReaderFrom).ReadFrom(r)
f.written += n
return n, err
}
func sanitizeMethod(m string) string {
switch m {
case "GET", "get":
return "get"
case "PUT", "put":
return "put"
case "HEAD", "head":
return "head"
case "POST", "post":
return "post"
case "DELETE", "delete":
return "delete"
case "CONNECT", "connect":
return "connect"
case "OPTIONS", "options":
return "options"
case "NOTIFY", "notify":
return "notify"
default:
return strings.ToLower(m)
}
}
func sanitizeCode(s int) string {
switch s {
case 100:
return "100"
case 101:
return "101"
case 200:
return "200"
case 201:
return "201"
case 202:
return "202"
case 203:
return "203"
case 204:
return "204"
case 205:
return "205"
case 206:
return "206"
case 300:
return "300"
case 301:
return "301"
case 302:
return "302"
case 304:
return "304"
case 305:
return "305"
case 307:
return "307"
case 400:
return "400"
case 401:
return "401"
case 402:
return "402"
case 403:
return "403"
case 404:
return "404"
case 405:
return "405"
case 406:
return "406"
case 407:
return "407"
case 408:
return "408"
case 409:
return "409"
case 410:
return "410"
case 411:
return "411"
case 412:
return "412"
case 413:
return "413"
case 414:
return "414"
case 415:
return "415"
case 416:
return "416"
case 417:
return "417"
case 418:
return "418"
case 500:
return "500"
case 501:
return "501"
case 502:
return "502"
case 503:
return "503"
case 504:
return "504"
case 505:
return "505"
case 428:
return "428"
case 429:
return "429"
case 431:
return "431"
case 511:
return "511"
default:
return strconv.Itoa(s)
}
}
// gzipAccepted returns whether the client will accept gzip-encoded content.
func gzipAccepted(header http.Header) bool {
a := header.Get(acceptEncodingHeader)
parts := strings.Split(a, ",")
for _, part := range parts {
part = strings.TrimSpace(part)
if part == "gzip" || strings.HasPrefix(part, "gzip;") {
return true
}
}
return false
}
// httpError removes any content-encoding header and then calls http.Error with
// the provided error and http.StatusInternalServerErrer. Error contents is
// supposed to be uncompressed plain text. However, same as with a plain
// http.Error, any header settings will be void if the header has already been
// sent. The error message will still be written to the writer, but it will
// probably be of limited use.
func httpError(rsp http.ResponseWriter, err error) {
rsp.Header().Del(contentEncodingHeader)
http.Error(
rsp,
"An error has occurred while serving metrics:\n\n"+err.Error(),
http.StatusInternalServerError,
)
}