dragonflydb / dragonflydb/dragonfly
large gap between used_memory_rss_human and used_memory on v1.36
- Dominant language
- C++
- Stars
- 31.5k
- Forks
- 1.3k
- Avg merge
- 1d 10h
- Merged PRs (30d)
- 137
Description
**Describe the bug**
After we upgraded the dragonfly version from 1.26 to 1.36 we have observed that total memory consumption has been increased significantly on top of the dragonfly configured memory, earlier the difference used to be 10-20gb which included OS as well but now this difference is around 50-100gb where total memory consumption is reaching the maximum memory provided by the instance.
**To Reproduce**
We only upgraded the dragonfly version from 1.26 to 1.36
**Expected behavior**
The difference between instance memory consumption and dragonfly memory consumption should have stayed the same as it use to be i.e 10-20gb.
**Screenshots**
**Environment (please complete the following information):**
- OS: [ubuntu 24.04]
- Kernel: # Command: `uname -a`
- Containerized?: Bare Metal
- Dragonfly Version: [1.36]
**Additional context**
We were not able to figure out how this instance memory consumption is reaching its maximum level even though the dragonfly memory consumption is following the same pattern as before.
Contributor guide
Research direction
Start by reproducing the upgrade from Dragonfly 1.26 to 1.36 on Ubuntu 24.04 bare metal and compare used_memory_rss_human with used_memory while tracking total instance memory. Review the 31-comment discussion for existing findings; done means the regression is isolated and a verified fix or clear root cause explains the additional memory.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- databases, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100