Please find attached sample JMeterParser.java, using XMLEncoder/XMLDecoder to save/retreive PerformanceReport to/from cached file.

Got problem with my git access so I couldn't provide a patch right now

Forget my attached JMeterParser, more works is required, especially in UriReport and PerformanceReport.

Some of my JMeter scripts (.jtl) could have more than 100000 lines and I've often between 30 and 60 reports historics.
Imagine all the memory allocated in Jenkins to get them all.
UriReport shouldn't be loaded with all HttpSamples but just counters (ie Average/90Line/...) of all samples).

Is there any progress on this issue?

This is still an issue and pretty much makes this plugin unusable for our purposes.

I'd suggest to use some kind of database to store all HttpSamples and let it do all the aggregation after the inital parsing of the jtl file.

It seems like no one is actually still working on this plugin though, which is kinda sad.

I believe that the fix for JENKINS-26975 will also solve this issue.

I would like to invite you to try the fix, which is available in this pull request: https://github.com/jenkinsci/performance-plugin/pull/44

The automated pull request handler generated a new plugin here (this artifact might become unavailable in the future though): https://jenkins.ci.cloudbees.com/job/plugins/job/performance-plugin/78/org.jenkins-ci.plugins$performance/

Plugin performance is fairly good even with large files now, using the latest version of the plugin. Reopen if this is still an issue. Please provide sample files to help reproduce the issue.