Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Hard-coded database connection pool size too small, requests failing #146

Open
abought opened this issue Mar 27, 2024 · 1 comment
Open
Labels
bug Something isn't working

Comments

@abought
Copy link
Collaborator

abought commented Mar 27, 2024

Problem

During periods of high traffic, users are randomly unable to connect to the website. Logs trace this to a connection pooling problem: although our database can support hundreds of simultaneous connections (our default value: 304), the pool size is capped at 10.

This results in elevated 500 errors, interrupted job uploads, and generally broken stuff. As the site gets more traffic, this becomes more common.

Proposed solution(s)

  1. Make maxActive connection pool size configurable when the database connection is first opened
  2. ...or at least increase the default value (based on metrics, 20-50 open connections would be sufficient, -1 for unlimited)
  3. (fixed, release is pending) Database indexing PR should reduce the time spent on each request, allowing open connections to be re-used sooner
  4. TIS will be increasing the time between "job status" refresh from 5 -> 20 sec in the next release. This should cut the load from our most common server request by ~75%.

Root cause

Many types of API endpoint require a database connection. A small pool of connections is re-used across all requests. If the pool is full, a request will time out without having a database connection. Ours logs have many stacktraces like the following; the error can be traced to the maxActive setting of dbcp. (renamed to maxTotal in newer versions)

2024-03-08 23:10:06,057 [qtp911501858-581359] ERROR cloudgene.mapred.database.DownloadDao - update download failed.
org.apache.commons.dbcp.SQLNestedException: Cannot get a connection, pool error Timeout waiting for idle object
at org.apache.commons.dbcp.PoolingDataSource.getConnection(PoolingDataSource.java:114) ~[commons-dbcp-1.4.jar:1.4]
at org.apache.commons.dbcp.BasicDataSource.getConnection(BasicDataSource.java:1044) ~[commons-dbcp-1.4.jar:1.4]
at org.apache.commons.dbutils.AbstractQueryRunner.prepareConnection(AbstractQueryRunner.java:319) ~[commons-dbutils-1.7.jar:1.7]
at org.apache.commons.dbutils.QueryRunner.update(QueryRunner.java:495) ~[commons-dbutils-1.7.jar:1.7]
at cloudgene.mapred.database.util.JdbcDataAccessObject.update(JdbcDataAccessObject.java:93) ~[cloudgene.jar:?]
at cloudgene.mapred.database.DownloadDao.update(DownloadDao.java:63) ~[cloudgene.jar:?]
at cloudgene.mapred.api.v2.jobs.ShareResults.get(ShareResults.java:48) ~[cloudgene.jar:?]
at sun.reflect.GeneratedMethodAccessor210.invoke(Unknown Source) ~[?:?]
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_402]
at java.lang.reflect.Method.invoke(Method.java:498) ~[?:1.8.0_402]
at org.restlet.resource.ServerResource.doHandle(ServerResource.java:511) ~[org.restlet-2.3.12.jar:?]
at org.restlet.resource.ServerResource.get(ServerResource.java:723) ~[org.restlet-2.3.12.jar:?]
at org.restlet.resource.ServerResource.doHandle(ServerResource.java:603) ~[org.restlet-2.3.12.jar:?]
at org.restlet.resource.ServerResource.doNegotiatedHandle(ServerResource.java:662) ~[org.restlet-2.3.12.jar:?]
at org.restlet.resource.ServerResource.doConditionalHandle(ServerResource.java:348) ~[org.restlet-2.3.12.jar:?]
at org.restlet.resource.ServerResource.handle(ServerResource.java:1020) ~[org.restlet-2.3.12.jar:?]
at org.restlet.resource.Finder.handle(Finder.java:236) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Router.doHandle(Router.java:422) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Router.handle(Router.java:641) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.application.StatusFilter.doHandle(StatusFilter.java:140) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.CompositeHelper.handle(CompositeHelper.java:202) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.application.ApplicationHelper.handle(ApplicationHelper.java:77) ~[org.restlet-2.3.12.jar:?]
at org.restlet.Application.handle(Application.java:385) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Router.doHandle(Router.java:422) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Router.handle(Router.java:641) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Router.doHandle(Router.java:422) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Router.handle(Router.java:641) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.application.StatusFilter.doHandle(StatusFilter.java:140) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.doHandle(Filter.java:150) ~[org.restlet-2.3.12.jar:?]
at org.restlet.routing.Filter.handle(Filter.java:197) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.CompositeHelper.handle(CompositeHelper.java:202) ~[org.restlet-2.3.12.jar:?]
at org.restlet.Component.handle(Component.java:408) ~[org.restlet-2.3.12.jar:?]
at org.restlet.Server.handle(Server.java:507) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.connector.ServerHelper.handle(ServerHelper.java:63) ~[org.restlet-2.3.12.jar:?]
at org.restlet.engine.adapter.HttpServerHelper.handle(HttpServerHelper.java:143) ~[org.restlet-2.3.12.jar:?]
at org.restlet.ext.jetty.JettyServerHelper$WrappedServer.handle(JettyServerHelper.java:256) ~[org.restlet.ext.jetty-2.3.12.jar:?]
at org.eclipse.jetty.server.HttpChannel.handle(HttpChannel.java:311) ~[jetty-server-9.2.14.v20151106.jar:9.2.14.v20151106]
at org.eclipse.jetty.server.HttpConnection.onFillable(HttpConnection.java:257) ~[jetty-server-9.2.14.v20151106.jar:9.2.14.v20151106]
at org.eclipse.jetty.io.AbstractConnection$2.run(AbstractConnection.java:544) ~[jetty-io-9.2.14.v20151106.jar:9.2.14.v20151106]
at org.eclipse.jetty.util.thread.QueuedThreadPool.runJob(QueuedThreadPool.java:635) ~[jetty-util-9.2.14.v20151106.jar:9.2.14.v20151106]
at org.eclipse.jetty.util.thread.QueuedThreadPool$3.run(QueuedThreadPool.java:555) ~[jetty-util-9.2.14.v20151106.jar:9.2.14.v20151106]
at java.lang.Thread.run(Thread.java:750) ~[?:1.8.0_402]
Caused by: java.util.NoSuchElementException: Timeout waiting for idle object
at org.apache.commons.pool.impl.GenericObjectPool.borrowObject(GenericObjectPool.java:1134) ~[commons-pool-1.5.4.jar:1.5.4]
at org.apache.commons.dbcp.PoolingDataSource.getConnection(PoolingDataSource.java:106) ~[commons-dbcp-1.4.jar:1.4]
... 57 more

@abought
Copy link
Collaborator Author

abought commented Apr 19, 2024

UM fork will be piloting maxActive = -1 in our next fork release, with server performance metrics to evaluate realistic load.

@abought abought added the bug Something isn't working label Apr 22, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

1 participant