Complete
Details
Assignee
Vladimir VinogradenkoVladimir VinogradenkoReporter
Ben SheetsBen SheetsLabels
Impact
HighOriginal estimate
Time remaining
0mComponents
Affects versions
Priority
Low
Details
Details
Assignee
Vladimir Vinogradenko
Vladimir VinogradenkoReporter
Ben Sheets
Ben SheetsLabels
Impact
High
Original estimate
Time remaining
0m
Components
Affects versions
Priority
Katalon Platform
Katalon Platform
Katalon Platform
Created July 2, 2022 at 3:17 PM
Updated February 27, 2025 at 10:31 PM
Resolved July 14, 2022 at 10:40 AM
Running TrueNAS-12.0-U8.1 with a remote graphite server configured. It's been working for a long time but I recently brought the graphite server down for an extended time.
While the graphite server has been down I've noticed the "Services" memory consumption of TrueNAS kept increasing and increasing. I confirmed with htop that it's collectd using more and more memory.
Collectd was up to 70% of system memory when I brought the graphite server online. Immediately after I brought graphite back online collectd started flushing it's cache. I watched it slowly flush all the way back down to ~1.5%.
I'm assuming an in memory cache is expected behavior but there should be a more sane limit to its size. It was up to 70% of my system's 32GB. I don't know if it would have caused problems when it got to 100% or if a limit would have kicked in.
Apologies if a high impact is inappropriate here but this felt like a memory leak to me.