Here is a non-exhaustive list of recommendations to get better performance:
- For VCS, check that common properties are set as prefetched (see org.nuxeo.ecm.core.schema.TypeService--configuration and org.nuxeo.ecm.core.schema.TypeService--doctype).
- For VCS, if you don't want to match proxies in your query, add a
AND ecm:isProxy = 0clause. - For VCS, if you don't use proxies at all, deactivate them at the repository level by adding inside the
<repository>element the following:<proxies enabled="false"/> - If you are doing a NXQL query that involves custom schemas you may need to add custom indexes to make the request efficient. You can trace the slow NXQL queries using NXP-14845.
- Use groups to manage ACLs. Adding a user to a group is free, but adding a user in an ACL at the document root level has a cost because optimized read ACLs need to be recomputed for all documents under the root.
- For VCS, when using Nuxeo in cluster mode, consider using Redis VCS row cache invalidation.
- For VCS, if possible disable the ACL optimization and switch all page providers to Elasticsearch. Look at the perf template.
- Disable listeners that are not used (see above perf template).
- Use BIGINT document identifier instead of default UID. This is done by adding a
<idType>sequence</idType>in the repository contribution). - Consider disabling the OS swapping (
sudo swapoff -a) or try to lower the swapiness (vm.swappiness=1). - Check the network latency between the application and the database.
- Configure ImageMagick to use a single thread.
- Monitor everything, JVM, GC, VCS cache hit ratio, database, system.