### What changes were proposed in this pull request?
This adds two new additional metrics to `ExternalBlockHandler`:
- `blockTransferRate` -- for indicating the rate of transferring blocks, vs. the data within them
- `blockTransferAvgSize_1min` -- a 1-minute trailing average of block sizes transferred by the ESS
Additionally, this enhances `YarnShuffleServiceMetrics` to expose the histogram/`Snapshot` information from `Timer` metrics within `ExternalBlockHandler`.
### Why are the changes needed?
Currently `ExternalBlockHandler` exposes some useful metrics, but is lacking around metrics for the rate of block transfers. We have `blockTransferRateBytes` to tell us the rate of _bytes_, but no metric to tell us the rate of _blocks_, which is especially relevant when running the ESS on HDDs that are sensitive to random reads. Many small block transfers can have a negative impact on performance, but won't show up as a spike in `blockTransferRateBytes` since the sizes are small. Thus the new metrics to show information around average block size and block transfer rate are very useful to monitor the health/performance of the ESS, especially when running on HDDs.
For the `YarnShuffleServiceMetrics`, currently the three `Timer` metrics exposed by `ExternalBlockHandler` are being underutilized in a YARN-based environment -- they are basically treated as a `Meter`, only exposing rate-based information, when the metrics themselves are collected detailed histograms of timing information. We should expose this information for better observability.
### Does this PR introduce _any_ user-facing change?
Yes, there are two entirely new metrics for the ESS, as documented in `monitoring.md`. Additionally in a YARN environment, `Timer` metrics exposed by the ESS will include more rich timing information.
### How was this patch tested?
New unit tests are added to verify that new metrics are showing up as expected.
We have been running this patch internally for approx. 1 year and have found it to be useful for monitoring the health of ESS and diagnosing performance issues.
Closes#32388 from xkrogen/xkrogen-SPARK-35258-ess-new-metrics.
Authored-by: Erik Krogen <xkrogen@apache.org>
Signed-off-by: Mridul Muralidharan <mridul<at>gmail.com>
### What changes were proposed in this pull request?
This PR upgrades Dropwizard metrics to 4.2.0.
I also modified the corresponding links in `docs/monitoring.md`.
### Why are the changes needed?
The latest version was released last week and it contains some improvements.
https://github.com/dropwizard/metrics/releases/tag/v4.2.0
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Build succeeds and all the modified links are reachable.
Closes#32628 from sarutak/upgrade-dropwizard-4.2.0.
Authored-by: Kousuke Saruta <sarutak@oss.nttdata.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Add toc tag on monitoring.md
### Why are the changes needed?
fix doc
### Does this PR introduce _any_ user-facing change?
yes, the table of content of the monitoring page will be shown on the official doc site.
### How was this patch tested?
pass GA doc build
Closes#32545 from yaooqinn/minor.
Authored-by: Kent Yao <yao@apache.org>
Signed-off-by: Kent Yao <yao@apache.org>
### What changes were proposed in this pull request?
This proposes to document the available metrics for ExecutorAllocationManager in the Spark monitoring documentation.
### Why are the changes needed?
The ExecutorAllocationManager is instrumented with metrics using the Spark metrics system.
The relevant work is in SPARK-7007 and SPARK-33763
ExecutorAllocationManager metrics are currently undocumented.
### Does this PR introduce _any_ user-facing change?
This PR adds documentation only.
### How was this patch tested?
na
Closes#32500 from LucaCanali/followupMetricsDocSPARK33763.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
Add more flexable parameters for stage end point
endpoint /application/{app-id}/stages. It can be:
/application/{app-id}/stages?details=[true|false]&status=[ACTIVE|COMPLETE|FAILED|PENDING|SKIPPED]&withSummaries=[true|false]$quantiles=[comma separated quantiles string]&taskStatus=[RUNNING|SUCCESS|FAILED|PENDING]
where
```
query parameter details=true is to show the detailed task information within each stage. The default value is details=false;
query parameter status can select those stages with the specified status. When status parameter is not specified, a list of all stages are generated.
query parameter withSummaries=true is to show both task summary information in percentile distribution and executor summary information in percentile distribution. The default value is withSummaries=false.
query parameter quantiles support user defined quantiles, default quantiles is `0.0,0.25,0.5,0.75,1.0`
query parameter taskStatus is to show only those tasks with the specified status within their corresponding stages. This parameter will be set when details=true (i.e. this parameter will be ignored when details=false).
```
### Why are the changes needed?
More flexable restful API
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
UT
Closes#31204 from AngersZhuuuu/SPARK-26399-NEW.
Lead-authored-by: Angerszhuuuu <angers.zhu@gmail.com>
Co-authored-by: AngersZhuuuu <angers.zhu@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Fix code not close issue in monitoring.md
### Why are the changes needed?
Fix doc issue
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Not need
Closes#32008 from AngersZhuuuu/SPARK-34911.
Authored-by: Angerszhuuuu <angers.zhu@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
When we want to get stage's detail info with task information, it will return all tasks, the content is huge and always we just want to know some failed tasks/running tasks with whole stage info to judge is a task has some problem. This pr support
user to use
```
/application/[appid]/stages/[stage-id]?details=true&taskStatus=xxx
/application/[appid]/stages/[stage-id]/[stage-attempted-id]?details=true&taskStatus=xxx
```
to filter task details by task status
### Why are the changes needed?
More flexiable Restful API
### Does this PR introduce _any_ user-facing change?
User can use
```
/application/[appid]/stages/[stage-id]?details=true&taskStatus=xxx
/application/[appid]/stages/[stage-id]/[stage-attempted-id]?details=true&taskStatus=xxx
```
to filter task details by task status
### How was this patch tested?
Added
Closes#31165 from AngersZhuuuu/SPARK-34092.
Lead-authored-by: Angerszhuuuu <angers.zhu@gmail.com>
Co-authored-by: angerszhu <angers.zhu@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
For a specific stage, it is useful to show the task metrics in percentile distribution. This information can help users know whether or not there is a skew/bottleneck among tasks in a given stage. We list an example in taskMetricsDistributions.json
Similarly, it is useful to show the executor metrics in percentile distribution for a specific stage. This information can show whether or not there is a skewed load on some executors. We list an example in executorMetricsDistributions.json
We define `withSummaries` and `quantiles` query parameter in the REST API for a specific stage as:
applications/<application_id>/<application_attempt/stages/<stage_id>/<stage_attempt>?withSummaries=[true|false]& quantiles=0.05,0.25,0.5,0.75,0.95
1. withSummaries: default is false, define whether to show current stage's taskMetricsDistribution and executorMetricsDistribution
2. quantiles: default is `0.0,0.25,0.5,0.75,1.0` only effect when `withSummaries=true`, it define the quantiles we use when calculating metrics distributions.
When withSummaries=true, both task metrics in percentile distribution and executor metrics in percentile distribution are included in the REST API output. The default value of withSummaries is false, i.e. no metrics percentile distribution will be included in the REST API output.
### Why are the changes needed?
For a specific stage, it is useful to show the task metrics in percentile distribution. This information can help users know whether or not there is a skew/bottleneck among tasks in a given stage. We list an example in taskMetricsDistributions.json
### Does this PR introduce _any_ user-facing change?
User can use below restful API to get task metrics distribution and executor metrics distribution for indivial stage
```
applications/<application_id>/<application_attempt/stages/<stage_id>/<stage_attempt>?withSummaries=[true|false]
```
### How was this patch tested?
Added UT
Closes#31611 from AngersZhuuuu/SPARK-34488.
Authored-by: Angerszhuuuu <angers.zhu@gmail.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
The endpoints of Prometheus metrics are properly marked and documented as an experimental (SPARK-31674). The class `PrometheusServlet` itself is not the part of an API so this PR proposes to remove it.
### Why are the changes needed?
To avoid marking a non-API as an API.
### Does this PR introduce _any_ user-facing change?
No, the class is already `private[spark]`.
### How was this patch tested?
Existing tests should cover.
Closes#31640 from HyukjinKwon/SPARK-34531.
Lead-authored-by: HyukjinKwon <gurwls223@apache.org>
Co-authored-by: Dongjoon Hyun <dongjoon@apache.org>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Removed the **hiveClientCalls.count** in CodeGenerator metrics in Component instance = Executor
### Why are the changes needed?
Wrong information regarding metrics was being displayed on Monitoring Documentation. I had added referred documentation for adding metrics logging in Graphite. This metric was not being reported. I had to check if the issue was at my application end or spark code or documentation. Documentation had the wrong info.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Manual, checked it on my forked repository feature branch [SPARK-33942](https://github.com/coderbond007/spark/blob/SPARK-33942/docs/monitoring.md)
Closes#30976 from coderbond007/SPARK-33942.
Authored-by: Pradyumn Agrawal (pradyumn.ag) <pradyumn.ag@media.net>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
This PR intends to fix typos in the sub-modules:
* `bin`
* `core`
* `docs`
* `external`
* `mllib`
* `repl`
* `pom.xml`
Split per srowen https://github.com/apache/spark/pull/30323#issuecomment-728981618
NOTE: The misspellings have been reported at 706a726f87 (commitcomment-44064356)
### Why are the changes needed?
Misspelled words make it harder to read / understand content.
### Does this PR introduce _any_ user-facing change?
There are various fixes to documentation, etc...
### How was this patch tested?
No testing was performed
Closes#30530 from jsoref/spelling-bin-core-docs-external-mllib-repl.
Authored-by: Josh Soref <jsoref@users.noreply.github.com>
Signed-off-by: Takeshi Yamamuro <yamamuro@apache.org>
### What changes were proposed in this pull request?
This PR aims to generalize executor metrics to support user-given file system schemes instead of the fixed `file,hdfs` scheme.
### Why are the changes needed?
For the users using only cloud storages like `S3A`, we need to be able to expose `S3A` metrics. Also, we can skip unused `hdfs` metrics.
### Does this PR introduce _any_ user-facing change?
Yes, but compatible for the existing users which uses `hdfs` and `file` filesystem scheme only.
### How was this patch tested?
Manually do the following.
```
$ build/sbt -Phadoop-cloud package
$ sbin/start-master.sh; sbin/start-slave.sh spark://$(hostname):7077
$ bin/spark-shell --master spark://$(hostname):7077 -c spark.executor.metrics.fileSystemSchemes=file,s3a -c spark.metrics.conf.executor.sink.jmx.class=org.apache.spark.metrics.sink.JmxSink
scala> spark.read.textFile("s3a://dongjoon/README.md").collect()
```
Separately, launch `jconsole` and check `*.executor.filesystem.s3a.*`. Also, confirm that there is no `*.executor.filesystem.hdfs.*`
```
$ jconsole
```
Closes#30405 from dongjoon-hyun/SPARK-33476.
Authored-by: Dongjoon Hyun <dongjoon@apache.org>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
This PR proposes to register the executor source with the Spark metrics system when running in local mode.
### Why are the changes needed?
The Apache Spark metrics system provides many useful insights on the Spark workload.
In particular, the [executor source metrics](https://github.com/apache/spark/blob/master/docs/monitoring.md#component-instance--executor) provide detailed info, including the number of active tasks, I/O metrics, and several task metrics details. The executor source metrics, contrary to other sources (for example ExecutorMetrics source), is not available when running in local mode.
Having executor metrics in local mode can be useful when testing and troubleshooting Spark workloads in a development environment. The metrics can be fed to a dashboard to see the evolution of resource usage and can be used to troubleshoot performance,
as [in this example](https://github.com/cerndb/spark-dashboard).
Currently users will have to deploy on a cluster to be able to collect executor source metrics, while the possibility of having them in local mode is handy for testing.
### Does this PR introduce _any_ user-facing change?
- This PR exposes executor source metrics data when running in local mode.
### How was this patch tested?
- Manually tested by running in local mode and inspecting the metrics listed in http://localhost:4040/metrics/json/
- Also added a test in `SourceConfigSuite`
Closes#28528 from LucaCanali/metricsWithLocalMode.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Thomas Graves <tgraves@apache.org>
### What changes were proposed in this pull request?
this PR renames the blacklisting feature. I ended up using "excludeOnFailure" or "excluded" in most cases but there is a mix. I renamed the BlacklistTracker to HealthTracker, but for the TaskSetBlacklist HealthTracker didn't make sense to me since its not the health of the taskset itself but rather tracking the things its excluded on so I renamed it to be TaskSetExcludeList. Everything else I tried to use the context and in most cases excluded made sense. It made more sense to me then blocked since you are basically excluding those executors and nodes from scheduling tasks on them. Then can be unexcluded later after timeouts and such. The configs I changed the name to use excludeOnFailure which I thought explained it.
I unfortunately couldn't get rid of some of them because its part of the event listener and history files. To keep backwards compatibility I kept the events and some of the parsing so that the history server would still properly read older history files. It is not forward compatible though - meaning a new application write the "Excluded" events so the older history server won't properly read display them as being blacklisted.
A few of the files below are showing up as deleted and recreated even though I did a git mv on them. I'm not sure why.
### Why are the changes needed?
get rid of problematic language
### Does this PR introduce _any_ user-facing change?
Config name changes but the old configs still work but are deprecated.
### How was this patch tested?
updated tests and also manually tested the UI changes and manually tested the history server reading older versions of history files and vice versa.
Closes#29906 from tgravescs/SPARK-32037.
Lead-authored-by: Thomas Graves <tgraves@nvidia.com>
Co-authored-by: Thomas Graves <tgraves@apache.org>
Signed-off-by: Thomas Graves <tgraves@apache.org>
### What changes were proposed in this pull request?
Update monitoring doc following the improvement/fix in SPARK-32119.
### Why are the changes needed?
SPARK-32119 removes the limitations listed in the monitoring doc "Distribution of the jar files containing the plugin code is currently not done by Spark."
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Not relevant
Closes#29463 from LucaCanali/followupSPARK32119.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Kousuke Saruta <sarutak@oss.nttdata.com>
### What changes were proposed in this pull request?
This PR fixes the link to metrics.dropwizard.io in monitoring.md to refer the proper version of the library.
### Why are the changes needed?
There are links to metrics.dropwizard.io in monitoring.md but the link targets refer the version 3.1.0, while we use 4.1.1.
Now that users can create their own metrics using the dropwizard library, it's better to fix the links to refer the proper version.
### Does this PR introduce _any_ user-facing change?
Yes. The modified links refer the version 4.1.1.
### How was this patch tested?
Build the docs and visit all the modified links.
Closes#29426 from sarutak/fix-dropwizard-url.
Authored-by: Kousuke Saruta <sarutak@oss.nttdata.com>
Signed-off-by: Sean Owen <srowen@gmail.com>
### What changes were proposed in this pull request?
Document the dependency between the config `spark.metrics.staticSources.enabled` and JVMSource registration.
### Why are the changes needed?
This PT just documents the dependency between config `spark.metrics.staticSources.enabled` and JVM source registration.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Manually tested.
Closes#29203 from LucaCanali/bugJVMMetricsRegistration.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
Support fetching taskList by status as below:
```
/applications/[app-id]/stages/[stage-id]/[stage-attempt-id]/taskList?status=failed
```
### Why are the changes needed?
When there're large number of tasks in one stage, current api is hard to get taskList by status
### Does this PR introduce _any_ user-facing change?
Yes. Updated monitoring doc.
### How was this patch tested?
Added tests in `HistoryServerSuite`
Closes#28942 from warrenzhu25/SPARK-32125.
Authored-by: Warren Zhu <zhonzh@microsoft.com>
Signed-off-by: Gengliang Wang <gengliang.wang@databricks.com>
### What changes were proposed in this pull request?
Add a new class HybridStore to make the history server faster when loading event files. When rebuilding the application state from event logs, HybridStore will write data to InMemoryStore at first and use a background thread to dump data to LevelDB once the writing to InMemoryStore is completed. HybridStore is to make content serving faster by using more memory. It's only safe to enable it when the cluster is not having a heavy load.
### Why are the changes needed?
HybridStore can greatly reduce the event logs loading time, especially for large log files. In general, it has 4x - 6x UI loading speed improvement for large log files. The detailed result is shown in comments.
### Does this PR introduce any user-facing change?
This PR adds new configs `spark.history.store.hybridStore.enabled` and `spark.history.store.hybridStore.maxMemoryUsage`.
### How was this patch tested?
A test suite for HybridStore is added. I also manually tested it on 3.1.0 on mac os.
This is a follow-up for the work done by Hieu Huynh in 2019.
Closes#28412 from baohe-zhang/SPARK-31608.
Authored-by: Baohe Zhang <baohe.zhang@verizonmedia.com>
Signed-off-by: Jungtaek Lim (HeartSaVioR) <kabhwan.opensource@gmail.com>
### What changes were proposed in this pull request?
SQL Rest API exposes query execution details and metrics as Public API. Its documentation will be useful for the end-users.
### Why are the changes needed?
SQL Rest API does not exist under Spark Rest API.
### Does this PR introduce any user-facing change?
No
### How was this patch tested?
Manually build and check
Closes#28354 from erenavsarogullari/SPARK-31566.
Lead-authored-by: Eren Avsarogullari <eren.avsarogullari@gmail.com>
Co-authored-by: Gengliang Wang <gengliang.wang@databricks.com>
Co-authored-by: Eren Avsarogullari <erenavsarogullari@gmail.com>
Signed-off-by: Gengliang Wang <gengliang.wang@databricks.com>
### What changes were proposed in this pull request?
This PR aims to new Prometheus-format metric endpoints experimental in Apache Spark 3.0.0.
### Why are the changes needed?
Although the new metrics are disabled by default, we had better make it experimental explicitly in Apache Spark 3.0.0 since the output format is still not fixed. We can finalize it in Apache Spark 3.1.0.
### Does this PR introduce _any_ user-facing change?
Only doc-change is visible to the users.
### How was this patch tested?
Manually check the code since this is a documentation and class annotation change.
Closes#28495 from dongjoon-hyun/SPARK-31674.
Authored-by: Dongjoon Hyun <dongjoon@apache.org>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
This PR proposes to remove the non-existed `hiveClientCalls.count` metric documentation of `CodeGenerator` of the Spark metrics system in the monitoring guide.
There is a duplicated `hiveClientCalls.count` metric in both `namespace=HiveExternalCatalog` and `namespace=CodeGenerator` bullet lists, but there is only one defined inside object `HiveCatalogMetrics`.
Closes#28292 from wezhang/monitoringdoc.
Authored-by: Wei Zhang <wezhang@outlook.com>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
This PR (SPARK-30775) aims to improve the description of the executor metrics in the monitoring documentation.
### Why are the changes needed?
Improve and clarify monitoring documentation by:
- adding reference to the Prometheus end point, as implemented in [SPARK-29064]
- extending the list and descripion of executor metrics, following up from [SPARK-27157]
### Does this PR introduce any user-facing change?
Documentation update.
### How was this patch tested?
n.a.
Closes#27526 from LucaCanali/docPrometheusMetricsFollowupSpark29064.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
### What changes were proposed in this pull request?
This is a FOLLOW-UP PR for review comment on #27208 : https://github.com/apache/spark/pull/27208#pullrequestreview-347451714
This PR documents a new feature `Eventlog Compaction` into the new section of `monitoring.md`, as it only has one configuration on the SHS side and it's hard to explain everything on the description on the single configuration.
### Why are the changes needed?
Event log compaction lacks the documentation for what it is and how it helps. This PR will explain it.
### Does this PR introduce any user-facing change?
No.
### How was this patch tested?
Built docs via jekyll.
> change on the new section
<img width="951" alt="Screen Shot 2020-02-16 at 2 23 18 PM" src="https://user-images.githubusercontent.com/1317309/74599587-eb9efa80-50c7-11ea-942c-f7744268e40b.png">
> change on the table
<img width="1126" alt="Screen Shot 2020-01-30 at 5 08 12 PM" src="https://user-images.githubusercontent.com/1317309/73431190-2e9c6680-4383-11ea-8ce0-815f10917ddd.png">
Closes#27398 from HeartSaVioR/SPARK-30481-FOLLOWUP-document-new-feature.
Authored-by: Jungtaek Lim (HeartSaVioR) <kabhwan.opensource@gmail.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
Add a section to the Configuration page to document configurations for executor metrics.
At the same time, rename spark.eventLog.logStageExecutorProcessTreeMetrics.enabled to spark.executor.processTreeMetrics.enabled and make it independent of spark.eventLog.logStageExecutorMetrics.enabled.
### Why are the changes needed?
Executor metrics are new in Spark 3.0. They lack documentation.
Memory metrics as a whole are always collected, but the ones obtained from the process tree have to be optionally enabled. Making this depend on a single configuration makes for more intuitive behavior. Given this, the configuration property is renamed to better reflect its meaning.
### Does this PR introduce any user-facing change?
Yes, only in that the configurations are all new to 3.0.
### How was this patch tested?
Not necessary.
Closes#27329 from wypoon/SPARK-27324.
Authored-by: Wing Yew Poon <wypoon@cloudera.com>
Signed-off-by: Imran Rashid <irashid@cloudera.com>
### What changes were proposed in this pull request?
This patch addresses remaining functionality on event log compaction: integrate compaction into FsHistoryProvider.
This patch is next task of SPARK-30479 (#27164), please refer the description of PR #27085 to see overall rationalization of this patch.
### Why are the changes needed?
One of major goal of SPARK-28594 is to prevent the event logs to become too huge, and SPARK-29779 achieves the goal. We've got another approach in prior, but the old approach required models in both KVStore and live entities to guarantee compatibility, while they're not designed to do so.
### Does this PR introduce any user-facing change?
No.
### How was this patch tested?
Added UT.
Closes#27208 from HeartSaVioR/SPARK-30481.
Authored-by: Jungtaek Lim (HeartSaVioR) <kabhwan.opensource@gmail.com>
Signed-off-by: Marcelo Vanzin <vanzin@apache.org>
### What changes were proposed in this pull request?
Fixed typo in `docs` directory and in other directories
1. Find typo in `docs` and apply fixes to files in all directories
2. Fix `the the` -> `the`
### Why are the changes needed?
Better readability of documents
### Does this PR introduce any user-facing change?
No
### How was this patch tested?
No test needed
Closes#26976 from kiszk/typo_20191221.
Authored-by: Kazuaki Ishizaki <ishizaki@jp.ibm.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
## What changes were proposed in this pull request?
This PR proposes to add instrumentation of memory usage via the Spark Dropwizard/Codahale metrics system. Memory usage metrics are available via the Executor metrics, recently implemented as detailed in https://issues.apache.org/jira/browse/SPARK-23206.
Additional notes: This takes advantage of the metrics poller introduced in #23767.
## Why are the changes needed?
Executor metrics bring have many useful insights on memory usage, in particular on the usage of storage memory and executor memory. This is useful for troubleshooting. Having the information in the metrics systems allows to add those metrics to Spark performance dashboards and study memory usage as a function of time, as in the example graph https://issues.apache.org/jira/secure/attachment/12962810/Example_dashboard_Spark_Memory_Metrics.PNG
## Does this PR introduce any user-facing change?
Adds `ExecutorMetrics` source to publish executor metrics via the Dropwizard metrics system. Details of the available metrics in docs/monitoring.md
Adds configuration parameter `spark.metrics.executormetrics.source.enabled`
## How was this patch tested?
Tested on YARN cluster and with an existing setup for a Spark dashboard based on InfluxDB and Grafana.
Closes#24132 from LucaCanali/memoryMetricsSource.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Imran Rashid <irashid@cloudera.com>
### What changes were proposed in this pull request?
This proposes to introduce a naming convention for Spark metrics configuration parameters used to enable/disable metrics source reporting using the Dropwizard metrics library: `spark.metrics.sourceNameCamelCase.enabled` and update 2 parameters to use this naming convention.
### Why are the changes needed?
Currently Spark has a few parameters to enable/disable metrics reporting. Their naming pattern is not uniform and this can create confusion. Currently we have:
`spark.metrics.static.sources.enabled`
`spark.app.status.metrics.enabled`
`spark.sql.streaming.metricsEnabled`
### Does this PR introduce any user-facing change?
Update parameters for enabling/disabling metrics reporting new in Spark 3.0: `spark.metrics.static.sources.enabled` -> `spark.metrics.staticSources.enabled`, `spark.app.status.metrics.enabled` -> `spark.metrics.appStatusSource.enabled`.
Note: `spark.sql.streaming.metricsEnabled` is left unchanged as it is already in use in Spark 2.x.
### How was this patch tested?
Manually tested
Closes#26692 from LucaCanali/uniformNamingMetricsEnableParameters.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
### What changes were proposed in this pull request?
The Spark metrics system produces many different metrics and not all of them are used at the same time. This proposes to introduce a configuration parameter to allow disabling the registration of metrics in the "static sources" category.
### Why are the changes needed?
This allows to reduce the load and clutter on the sink, in the cases when the metrics in question are not needed. The metrics registerd as "static sources" are under the namespaces CodeGenerator and HiveExternalCatalog and can produce a significant amount of data, as they are registered for the driver and executors.
### Does this PR introduce any user-facing change?
It introduces a new configuration parameter `spark.metrics.register.static.sources.enabled`
### How was this patch tested?
Manually tested.
```
$ cat conf/metrics.properties
*.sink.prometheusServlet.class=org.apache.spark.metrics.sink.PrometheusServlet
*.sink.prometheusServlet.path=/metrics/prometheus
master.sink.prometheusServlet.path=/metrics/master/prometheus
applications.sink.prometheusServlet.path=/metrics/applications/prometheus
$ bin/spark-shell
$ curl -s http://localhost:4040/metrics/prometheus/ | grep Hive
metrics_local_1573330115306_driver_HiveExternalCatalog_fileCacheHits_Count 0
metrics_local_1573330115306_driver_HiveExternalCatalog_filesDiscovered_Count 0
metrics_local_1573330115306_driver_HiveExternalCatalog_hiveClientCalls_Count 0
metrics_local_1573330115306_driver_HiveExternalCatalog_parallelListingJobCount_Count 0
metrics_local_1573330115306_driver_HiveExternalCatalog_partitionsFetched_Count 0
$ bin/spark-shell --conf spark.metrics.static.sources.enabled=false
$ curl -s http://localhost:4040/metrics/prometheus/ | grep Hive
```
Closes#26320 from LucaCanali/addConfigRegisterStaticMetrics.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
Spark 2.4 added the ability for executor plugins to be loaded into
Spark (see SPARK-24918). That feature intentionally skipped the
driver to keep changes small, and also because it is possible to
load code into the Spark driver using listeners + configuration.
But that is a bit awkward, because the listener interface does not
provide hooks into a lot of Spark functionality. This change reworks
the executor plugin interface to also extend to the driver.
- there's a "SparkPlugin" main interface that provides APIs to
load driver and executor components.
- custom metric support (added in SPARK-28091) can be used by
plugins to register metrics both in the driver process and in
executors.
- a communication channel now exists that allows the plugin's
executor components to send messages to the plugin's driver
component easily, using the existing Spark RPC system.
The latter was a feature intentionally left out of the original
plugin design (also because it didn't include a driver component).
To avoid polluting the "org.apache.spark" namespace, I added the new
interfaces to the "org.apache.spark.api" package, which seems like
a better place in any case. The actual implementation is kept in
an internal package.
The change includes unit tests for the new interface and features,
but I've also been running a custom plugin that extends the new
API in real applications.
Closes#26170 from vanzin/SPARK-29397.
Authored-by: Marcelo Vanzin <vanzin@cloudera.com>
Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com>
This adds an entry about PrometheusServlet to the documentation, following SPARK-29032
### Why are the changes needed?
The monitoring documentation lists all the available metrics sinks, this should be added to the list for completeness.
Closes#26081 from LucaCanali/FollowupSpark29032.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
## What changes were proposed in this pull request?
This proposes to improve Spark instrumentation by adding a hook for user-defined metrics, extending Spark’s Dropwizard/Codahale metrics system.
The original motivation of this work was to add instrumentation for S3 filesystem access metrics by Spark job. Currently, [[ExecutorSource]] instruments HDFS and local filesystem metrics. Rather than extending the code there, we proposes with this JIRA to add a metrics plugin system which is of more flexible and general use.
Context: The Spark metrics system provides a large variety of metrics, see also , useful to monitor and troubleshoot Spark workloads. A typical workflow is to sink the metrics to a storage system and build dashboards on top of that.
Highlights:
- The metric plugin system makes it easy to implement instrumentation for S3 access by Spark jobs.
- The metrics plugin system allows for easy extensions of how Spark collects HDFS-related workload metrics. This is currently done using the Hadoop Filesystem GetAllStatistics method, which is deprecated in recent versions of Hadoop. Recent versions of Hadoop Filesystem recommend using method GetGlobalStorageStatistics, which also provides several additional metrics. GetGlobalStorageStatistics is not available in Hadoop 2.7 (had been introduced in Hadoop 2.8). Using a metric plugin for Spark would allow an easy way to “opt in” using such new API calls for those deploying suitable Hadoop versions.
- We also have the use case of adding Hadoop filesystem monitoring for a custom Hadoop compliant filesystem in use in our organization (EOS using the XRootD protocol). The metrics plugin infrastructure makes this easy to do. Others may have similar use cases.
- More generally, this method makes it straightforward to plug in Filesystem and other metrics to the Spark monitoring system. Future work on plugin implementation can address extending monitoring to measure usage of external resources (OS, filesystem, network, accelerator cards, etc), that maybe would not normally be considered general enough for inclusion in Apache Spark code, but that can be nevertheless useful for specialized use cases, tests or troubleshooting.
Implementation:
The proposed implementation extends and modifies the work on Executor Plugin of SPARK-24918. Additionally, this is related to recent work on extending Spark executor metrics, such as SPARK-25228.
As discussed during the review, the implementaiton of this feature modifies the Developer API for Executor Plugins, such that the new version is incompatible with the original version in Spark 2.4.
## How was this patch tested?
This modifies existing tests for ExecutorPluginSuite to adapt them to the API changes. In addition, the new funtionality for registering pluginMetrics has been manually tested running Spark on YARN and K8S clusters, in particular for monitoring S3 and for extending HDFS instrumentation with the Hadoop Filesystem “GetGlobalStorageStatistics” metrics. Executor metric plugin example and code used for testing are available, for example at: https://github.com/cerndb/SparkExecutorPluginsCloses#24901 from LucaCanali/executorMetricsPlugin.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com>
## What changes were proposed in this pull request?
Today all registered metric sources are reported to GraphiteSink with no filtering mechanism, although the codahale project does support it.
GraphiteReporter (ScheduledReporter) from the codahale project requires you implement and supply the MetricFilter interface (there is only a single implementation by default in the codahale project, MetricFilter.ALL).
Propose to add an additional regex config to match and filter metrics to the GraphiteSink
## How was this patch tested?
Included a GraphiteSinkSuite that tests:
1. Absence of regex filter (existing default behavior maintained)
2. Presence of `regex=<regexexpr>` correctly filters metric keys
Closes#25232 from nkarpov/graphite_regex.
Authored-by: Nick Karpov <nick@nickkarpov.com>
Signed-off-by: jerryshao <jerryshao@tencent.com>
## What changes were proposed in this pull request?
The motivation for these additional metrics is to help in troubleshooting and monitoring task execution workload when running on a cluster. Currently available metrics include executor threadpool metrics for task completed and for active tasks. The addition of threadpool taskStarted metric will allow for example to collect info on the (approximate) number of failed tasks by computing the difference thread started – (active threads + completed tasks and/or successfully finished tasks).
The proposed metric finishedTasks is also intended for this type of troubleshooting. The difference between finshedTasks and threadpool.completeTasks, is that the latter is a (dropwizard library) gauge taken from the threadpool, while the former is a (dropwizard) counter computed in the [[Executor]] class, when a task successfully finishes, together with several other task metrics counters.
Note, there are similarities with some of the metrics introduced in SPARK-24398, however there are key differences, coming from the fact that this PR concerns the executor source, therefore providing metric values per executor + metric values do not require to pass through the listerner bus in this case.
## How was this patch tested?
Manually tested on a YARN cluster
Closes#22290 from LucaCanali/AddMetricExecutorStartedTasks.
Lead-authored-by: Luca Canali <luca.canali@cern.ch>
Co-authored-by: LucaCanali <luca.canali@cern.ch>
Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com>
## What changes were proposed in this pull request?
Up to now, Apache Spark maintains the given event log directory by **time** policy, `spark.history.fs.cleaner.maxAge`. However, there are two issues.
1. Some file system has a limitation on the maximum number of files in a single directory. For example, HDFS `dfs.namenode.fs-limits.max-directory-items` is 1024 * 1024 by default.
https://hadoop.apache.org/docs/r3.2.0/hadoop-project-dist/hadoop-hdfs/hdfs-default.xml
2. Spark is sometimes unable to to clean up some old log files due to permission issues (mainly, security policy).
To handle both (1) and (2), this PR aims to support an additional policy configuration for the maximum number of files in the event log directory, `spark.history.fs.cleaner.maxNum`. Spark will try to keep the number of files in the event log directory according to this policy.
## How was this patch tested?
Pass the Jenkins with a newly added test case.
Closes#25072 from dongjoon-hyun/SPARK-28294.
Authored-by: Dongjoon Hyun <dhyun@apple.com>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
## What changes were proposed in this pull request?
SPARK-27773 has introduced a new metric (counter) numCaughtExceptions to the Spark Dropwizard monitoring system. This PR adds an entry in the monitoring documentation to document this.
Closes#24790 from LucaCanali/addDocFollowingSPARK27773.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Dongjoon Hyun <dhyun@apple.com>
## What changes were proposed in this pull request?
Add AL2 license to metadata of all .md files.
This seemed to be the tidiest way as it will get ignored by .md renderers and other tools. Attempts to write them as markdown comments revealed that there is no such standard thing.
## How was this patch tested?
Doc build
Closes#24243 from srowen/SPARK-26918.
Authored-by: Sean Owen <sean.owen@databricks.com>
Signed-off-by: Sean Owen <sean.owen@databricks.com>
## What changes were proposed in this pull request?
This applies some minor updates/cleaning following up SPARK-26928, notably renaming JVMCPU.scala to JVMCPUSource.scala.
## How was this patch tested?
Manually tested
Closes#24201 from LucaCanali/fixupSPARK-26928.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Sean Owen <sean.owen@databricks.com>
## What changes were proposed in this pull request?
Add completedStages metircs for namespace=appStatus for monitoring.md.
Closes#24109 from hehuiyuan/hehuiyuan-patch-5.
Authored-by: hehuiyuan <hehuiyuan@ZBMAC-C02WD3K5H.local>
Signed-off-by: Sean Owen <sean.owen@databricks.com>
## What changes were proposed in this pull request?
A sub-task of [SPARK-23206](https://issues.apache.org/jira/browse/SPARK-23206)
Add Executor level metrics to monitoring docs
## How was this patch tested?
jekyll
Closes#24090 from LantaoJin/SPARK-27157.
Authored-by: Lantao Jin <jinlantao@gmail.com>
Signed-off-by: Sean Owen <sean.owen@databricks.com>
## What changes were proposed in this pull request?
This proposes to add instrumentation for the driver's JVM CPU time via the Spark Dropwizard/Codahale metrics system. It follows directly from previous work SPARK-25228 and shares similar motivations: it is intended as an improvement to be used for Spark performance dashboards and monitoring tools/instrumentation.
Implementation details: this PR takes the code introduced in SPARK-25228 and moves it to a new separate Source JVMCPUSource, which is then used to register the jvmCpuTime gauge metric for both executor and driver.
The registration of the jvmCpuTime metric for the driver is conditional, a new configuration parameter `spark.metrics.cpu.time.driver.enabled` (proposed default: false) is introduced for this purpose.
## How was this patch tested?
Manually tested, using local mode and using YARN.
Closes#23838 from LucaCanali/addCPUTimeMetricDriver.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com>
## What changes were proposed in this pull request?
This PR proposes to extend the documentation of the Spark metrics system in the monitoring guide. In particular by:
- adding a list of the available metrics grouped per component instance
- adding information on configuration parameters that can be used to configure the metrics system in alternative to the metrics.properties file
- adding information on the configuration parameters needed to enable certain metrics
- it also propose to add an example of Graphite sink configuration in metrics.properties.template
Closes#23798 from LucaCanali/metricsDocUpdate.
Authored-by: Luca Canali <luca.canali@cern.ch>
Signed-off-by: Sean Owen <sean.owen@databricks.com>
