Memory issue with

[Stack: OS: Windows; Java:OpenJDK 8u252-b09; Tomcat: 9.0.31]

Has anyone else had memory problems with the latest stable release

We upgraded one of our prod machines to it from ten days ago and it’s been fine.

But on our other one it seemed to result in a memory leak of some kind. This machine is quite busy but only during local business hours. We installed it on Sunday and all seemed well when quiet. But at 8am on Monday memory use started climbing abnormally.

Here’s what it normally looks like at that time on a weekday (running There is 8GB allocated which is more than adequate. You can see Garbage Collection running normally, and this is the general pattern for the rest of a typical day.

But this is what we saw with GC doesn’t seem to be recovering very much space:

I ran GC manually (using FR) but it didn’t have much effect either.

By mid afternoon it had climbed further:

…until by the end of the day the heap was full, the server ground to a halt and had to be restarted:

The same pattern played out again today more quickly so we’ve gone back to which has been solid for over 2 months (despite being a “snapshot”).

Perhaps it’s related to this issue, although this is definitely heap memory, not metaspace. Or maybe this one?

Doesn’t look like you have seen any response. We see something similar on our side on very busy servers. Used to GC every couple hours and regain 60% or memory. Now we see it very often and only gaining maybe 20% each time.

We are also running latest java8 and tomcat 9.0.31… Our Lucee version just jumped to from a 5.2 version that we don’t really care to bounce back to.

Thanks for chiming in with your experience, which confirms it only affects certain “busy” conditions. Unfortunately that makes it hard to identify the issue.

If you want to move on from 5.2, we’ve found to be very solid. Don’t be put off by the SNAPSHOT label.

Thanks for sharing this info Julian.

Anyone have input if anything is being addressed in a future release around this? Is bug created? I am trying to decide next steps and backing up a release may be something we pass to QA next week for review. We always just deploy released versions in production.

Julian, are you still running snapshot Or have you found a newer snapshot or release candidate?

Yes, seems like we’re stuck on this version for now. Can’t really keep testing newer releases as the only way of doing so is by applying them to this particular production server at busy times, which is obviously very disruptive if they fail.