forked from prometheus-net/prometheus-net
-
Notifications
You must be signed in to change notification settings - Fork 0
/
History
218 lines (218 loc) · 16.7 KB
/
History
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
* 8.2.1
- Fix occasional "Collection was modified" exception when serializing metrics. #464
* 8.2.0
- .WithLabels() & similar now accept ReadOnlyMemory<string> as alternative to string[]. Same behavior, just easier to use if you already have a ReadOnlyMemory<string>.
- .WithLabels() & similar now accept ReadOnlySpan<string> as alternative to string[]. This enables allocation-free metric instance creation if a metric instance with these labels is already known.
- Incorporated various optimizations to reduce the required CPU time and allocated memory, including #410, #443 and other contributions.
- Observation of large histograms is now 10-30% faster on machines that support AVX2 instructions.
- health checks exposed via ForwardToPrometheus() no longer create a default metric if a custom metric is provided #444
* 8.1.1
- Fix bug in .NET Meters API adapter for UpDownCounter, which was incorrectly transformed to Prometheus metrics. #452 and #453
* 8.1.0
- Add support for capturing HttpClient metrics from all registered HttpClients (`services.UseHttpClientMetrics()`).
* 8.0.1
- Allow ObservableCounter<T> to be reset. Previously, the MeterAdapter output got stuck on its previous maximum if the underlying Meter reset its value to a lower value. Now we transform the value 1:1.
* 8.0.0
- Added OpenMetrics exposition format support (#388).
- Added exemplar support for Counter and Histogram (#388).
- The ':' character is no longer allowed in metric or label names. For metric names, Prometheus standard practice is to use colon only in recording rules.
- Publish symbol packages and add Source Link support for easier debugging experience.
- Fix defect where metrics with different labels could overwrite each other in specific circumstances with multiple metric factories in use (#389).
- Ensure even harder that MetricPusher flushes the final state before stopping (#383 and #384)
- Simplify DotNetStats built-in collector code for ease of readability and more best practices (#365, #364)
- Slightly improve Counter performance under concurrent load.
- Reduce memory allocations performed during ASP.NET Core HTTP request tracking.
- By default, EventCounterAdapter will only listen to a small predefined set of general-purpose useful event sources, to minimize resource consumption in the default configuration. A custom event source filter must now be provided to enable listening for additional event sources.
- EventCounterAdapter will only refresh data every 10 seconds by default, to reduce amount of garbage generated in memory (.NET event counters are very noisy and create many temporary objects).
- Added `IManagedLifetimeMetricFactory.WithLabels()` to enable simpler label enrichment in scenarios where lifetime-managed metric instances are used.
* 7.1.0
- Added back .NET Standard 2.0 support as some customers had a hard dependency on .NET Standard 2.0 (despite not being a documented feature even earlier).
- Added (Observable)UpDownCounter support to MeterAdapter (.NET 7 specific feature).
* 7.0.0
- .NET Core specific functionality now targeting .NET 6.0 or greater (all previous versions will be end of life by December 2022).
- Relaxed the API restriction that forbade you to create multiple metrics with the same name but different label names. While this is a Prometheus anti-pattern, it is a standard pattern in other metrics technologies and we will allow it in the name of interoperability.
- Added IMetricFactory.WithLabels() to define static labels per-factory and reduce the effort needed by users when defining static labels in specific scopes.
- Added automatic metric expiration feature, to remove timeseries when they are no longer used after a while. This improves interoperability with other metrics technologies such as .NET Meters.
- (Breaking change) Rewrite of the MeterAdapter functionality to bridge .NET 6 metrics to Prometheus. API remains equivalent, though options class changed. Output format (what metrics are published) is completely different. Now it translates metric names (instead of putting them in labels) and preserves tags (labels) provided via the .NET Meters API. Added sample project to demonstrate functionality. Ref #359
- (Breaking change) Dropped support for .NET Standard 2.0 - now will target specifically .NET 6.0 and .NET Framework 4.6.2. This was already documented before but there was "unofficial" support also for .NET Standard remaining. This is now removed.
- (Breaking change) EventCounterAdapter output format changed to translate counters to Prometheus metrics and to improve usability by publishing both counters and gauges from ambiguous event counters.
- EventCounterAdapter and MeterAdapter are now enabled on startup by default.
- Added some debug metrics to indicate the number of metric families, instances and timeseries being exported by prometheus-net.
- Added typical benchmark results to readme for easy reference, and to show relative speed of the simple versus complex instruments.
- Removed MetricConfiguration.StaticLabels because it had negative performance impact on many code paths that did not use it. The same functionality can be manually implemented in user code. Static labels remain available on registry and factory level.
- When a before-collection metrics update callback throws an exception, it is now logged and ignored instead of breaking all metrics collection.
- EventCounterAdapter now defaults to Informational events only, for reduced performance overhead. Can be configured via callback.
* 6.0.0
- Dropped library support for unsupported .NET runtime versions (.NET 2.1).
- If Razor Pages is enabled, the "page" label is now automatically added to the default HTTP server metrics, unless there is a user-defined label already present with this name.
- The "endpoint" label is now automatically added to the default HTTP server metrics, containing this endpoint routing route pattern of the request, unless there is a user-defined label already present with this name.
- Added integration with .NET EventCounters, which can now be easily exposed as Promtheus metrics (see EventCounterAdapter class).
- Added integration with .NET Meters, which can now be easily exposed as Promtheus metrics (see MeterAdapter class).
- Added ReplaceOnPush option to MetricPusher #305 #322.
- Minimum .NET Framework version is now 4.6.2.
* 5.1.0
- Add support for adding custom label value providers to HTTP server metrics. #347
* 5.0.2
- Bugfix: HttpClientMetrics under concurrent load with long-running requests were leading to broken response streams. #336
* 5.0.1
- Bugfix: PowersOfTenDividedBuckets did not work with >= 10 buckets due to overlap. Now skips overlapping region of generated series.
* 5.0.0
- Added HTTP client metric that measures the response duration (time until response content reading is finished).
- Added "client" label containing the name of HTTP client to HTTP client metrics.
- Added "code" label with HTTP response code to HTTP client metrics. #299
- HTTP client metrics now count responses when they are completed, not when started, to match with HTTP server metrics. Also renamed metric to httpclient_requests_sent_total since it said "received" previously (due to copy-paste error from HTTP server metrics).
- HTTP client metrics are now published via the netstandard package, without requiring ASP.NET Core. #314
- Added Counter.IncTo[Current]TimeUtc().
* 4.2.0
- Added Histogram.PowersOfTenDividedBuckets()
- Improved accuracy of histogram bucket calculations to reduce the likelyhood of seeing 0.00000000000000000001 offsets from expected value due to floating point math.
- Added ConfigureAwait(false) to MetricHandler to help avoid deadlocks in some rare scenarios. #308
- Fix defect where MetricPusher terminated when HttpClient signaled timeout. #302, #293
- No longer need to manually reference ASP.NET Core SDK in csproj file when installing prometheus-net.AspNetCore library. It gets automatically referenced now. #267
* 4.1.1
- Bugfix: HTTP request metrics in 4.1.0 got broken due to typo in middleware helper.
* 4.1.0
- Fixed HttpClient metrics duration that was incorrectly measured. #265
- prometheus-net.NetFramework.AspNet is now strong named, just like all the other assemblies. #273
- NetFx target is now conditional for Windows builds only, to allow easier development work on nonwindows platforms. #264
- MetricPusher now supports a callback for logging push errors. #186
- Added support for defining static labels on registry or metric level. #256
- Added ICollectorRegistry and IMetricFactory for mocking in test code. #268
- Ignore /metrics URL by default in HTTP request metrics.
- Fix overflow in Summary after observing more than uint.MaxValue events. #277
- Reduce allocations in hot paths. #276, #275
- Added request filtering (authorization) support to standalone MetricServer. #261
* 4.0.0
- Allow arbitrary ASP.NET Core route parameters to be exposed as HTTP metric labels. #225
- Breaking change: some HTTP metrics implementation classes were changed from public to internal. They were never supposed to be public and hopefully nobody ever used them directly. Public documented API did not change.
- Add HttpClient metrics with IHttpClientFactory integration.
* 3.6.0
- Added ASP.NET Core health check publisher for publishing metrics on health check status. #203 #209
- Fixed hang in ASP.NET Web API exporter. #215
- Fixed ASP.NET Web API exporter exporting metrics from the wrong registry. #217
- Reduced memory use in MetricPusher implementation. Inspired by #208
- Added basic gRPC service metrics support. #212
- Reduce pointless log spam on cancelled scrapes - will silently ignore cancelled scrapes in the ASP.NET Core exporter. #220
* 3.5.0
- Exposed metric.Unpublish() method since there was already a matching Publish() there.
- When the scrape is aborted, stop collecting/serializing metrics. #189
- Added simple metric export handler for legacy ASP.NET Web API, to avoid having to deal with HttpListener complexity.
* 3.4.0
- Separate packaging for ASP.NET Core 3 to reduce package reference management complexity for users. #164
- Lazy-initialize the default sample metrics to reduce initialization cost for those who do not need them. #170
- ASP.NET Core middlware no longer disposes of output stream. #171
- Fix DotNetStats occasional data tearing under parallel collections. #173
- Improved wording of some error messages. #169
- Added Counter.IncTo(), Gauge.IncTo(), Gauge.DecTo(). #181
- Added Gauge.Set(DateTimeOffset) and Gauge.IncTo(DateTimeOffset). #181
- Improve .NET Framework 4.6.1 packaging to include the ValueTuple reference automatically. #179
- ASP.NET Core HTTP metrics now capture route data before it is overwritten by an exception handler. #177
- ASP.NET Core HTTP "in progress" metrics now include controller/action labels (ASP.NET Core 3 only).
- Fixed concurrently defect in HTTP metrics that may have caused label values from wrong request to apply. #182
- The HttpListener based MetricServer can now handle more than one concurrent request.
- Added asynchronous "before collect" callbacks, giving user code more flexibility in metric updates.
- Added .Remove() to labeled metric instances for more convenient removal API (previously had to call method on parent instance).
- Added .Unpublish() to labeled metric instances to support temporary unpublishing of a metric, without discarding collected data (e.g. histogram buckets), during periods of data unavailability.
* 3.3.0
- Add ability to filter requests by port number when exporting metrics. #160
* 3.2.1
- Fix NuGet packaging defect that caused reference failures after installing .2.0 on .NET Core 2.1 or newer. #158
* 3.2.0
- Enable custom HTTP client to be used in MetricPusher, allowing for Basic auth and more flexibility. #125
- Add DiagnosticSourceAdapter (.NET Core only). #153
- Improved compatibility with older .NET Framework versions. #155
* 3.1.4
- Fix timezone mismatch in Summary that led to unintended delays in buffer flushing. #149
* 3.1.3
- Fix regression introduced in 3.1.0 where MetricPusher does not always flush metrics before stopping.
* 3.1.2
- Fix defect where Histogram batch observations only incremented sum by one value, instead of entire batch. #147
* 3.1.1
- Added missing UTF-8 charset to Content-Type header, so non-ASCII characters are interpreted correctly by browsers.
* 3.1.0
- Added ICounter.NewTimer() (adds the value to the counter)
- Eliminated some needless allocations when trying to register a metric that is already registered. #134
- Added IHistogram.Count and IHistogram.Sum to expose aspects of collected data for inspection.
- Added Collector.GetAllLabelValues() to expose the list of child metrics by all their known label values.
- Metric export is now asynchronous internally to be compatible with ASP.NET Core 3.0 default configuration.
- Added CollectorRegistry.CollectAndExportAsTextAsync() to support metric data export via arbitrary custom endpoints.
* 3.0.3
- Now backward compatible with ASP.NET Core 2.1 (was 2.2+)
* 3.0.2
- Fix defect where histogram sum failed to increment.
* 3.0.1
- Fix ObjectDisposedException in MetricPusher.
* 3.0.0
- Added HTTP request metrics for ASP.NET Core.
- Somewhat more realistic examples in readme.
- Metrics exporter is now significantly more CPU and memory-efficient.
- Added Observe(value, count) to histogram metric, enabling multiple observations with the same value to be counted.
- Added CountExceptions() and MeasureInProgress() helper extensions.
- Adjusted API to better conform to Prometheus client library guidelines in terms of default values.
- Breaking change: assemblies are now strong-named.
- Breaking change: removed "windows" from built-in metric names as they are not Windows-specific.
- Breaking change: removed support for protobuf export format (it is no longer used by Prometheus).
- Breaking change: API surface cleaned up, removed some legacy methods, made many internal types actually internal.
- Breaking change: "on demand collectors" concept replaced with simpler "before collect callbacks". Works the same, just less code needed to use it and fewer possible error conditions.
- Breaking change: removed support for "custom collectors", as this was a very special use case that did not benefit at all from the main functionality of the library. Just generate a Prometheus exporter output document yourself if you need to export arbitrary data.
* 2.1.3
- Fixed wrong case used for metric type in the export data format. Should always be lowercase. #96
* 2.1.2
- Fixed potential conflict when using pushgateway and also other exporter libraries (see #89)
* 2.1.1
- Various minor fixes (see issues on GitHub for details).
* 2.1.0
- Add MetricOptions and subclasses for more extensible API (old API surface remains available)
- Add SuppressInitialValue to metric configuration (ref https://www.robustperception.io/existential-issues-with-metrics/)
- Add .WithLabels() as alternative to .Labels() for fewer annoying Intellisense conflicts.
* 2.0.0
- Targeting .NET Standard 2.0 as minimum version (.NET Framework 4.6.1, .NET Core 2.0 and Mono 5.4)
- Added ASP.NET Core middlware
- Added possibility to signal a failed scrape from on-demand collectors
- Removed dependency on Reactive Extensions
- Minor breaking changes to API
- Performance improvements for hot-path code
- Removed mostly obsolete PerfCounterCollector class
- Fixed NuGet package contents to remove assemblies from dependencies
- Various minor fixes (see issues on GitHub for details)
* 1.3.4
- Added support for .NET 4.5 using System.Reactive 3.1.1.
- .NET 4.0 support continues to target Rx 2.5
* 1.2.4:
- Fixed MetricPusher not flushing metrics when stopped
* 1.2.3:
- Fixed label values escaping for ASCII formatter
* 1.2.2:
- PushGateway support
- Various internal improvements (replaced locks with Interlocked operations)
* 1.1.4:
- Fixed some metrics not updating, added process ID metric
- Replaced lock statements in Counter and Gauge with CAS
* 1.1.3:
- optionally use https in MetricServer
* 1.1.2:
- using UTF-8 in text formatter
- catching exceptions in MetricServer http loop
* 1.1.1:
- disposing of MetricServer loop on Stop()
* 1.1.0:
- Renamed some metric names to be in-line with prometheus guidelines (breaking change as far as the exported metrics are concerned)
* 1.0.0:
- Add CPU, num handles, start time, num threads metrics to dot net stats collector
- Made DotNetStatsCollector default (previously it was PerfCounterCollector)
* 0.0.11:
- Summary metric ported from go
* 0.0.10:
- Fix header writing order
* 0.0.9:
- Generalise scraping so it can be called externally without using the embedded http handler
* 0.0.8:
- Introduced interfaces for all the metrics to make unlabelled collectors and their children polymorph
* 0.0.7:
- Added the notion of OnDemandCollectors + a DotNetStatsCollector to avoid having to use .net perf counters
* 0.0.6:
- Do not create unlabelled metric if label names are specified
* 0.0.5:
- Allow specifying hostname in URL
- Fix null ref exception if 'Accept' header is not specified
* 0.0.3 - initial version