Repository navigation
Closed as duplicate
Closed as duplicate
Copy link
Labels
area:corebugSomething isn't workingSomething isn't workinghas reproHas detailed reproduction stepsHas detailed reproduction stepsmemoryperf:memoryplatform:linuxIssue specifically occurs on LinuxIssue specifically occurs on Linux
Description
Claude Code v1.0.53 is frequently and apparently randomly using excessive amounts of memory and getting OOM killed.
analyze this dmesg
```[6746807.992310] [1661706] 1000 1661706 2945110 58453 9555968 0 0 node
[6746807.992312] [1708298] 0 1708298 696366 5064 471040 0 -500 dockerd
[6746807.992315] [1717450] 1000 1717450 8214417 30402 3702784 0 0 node
[6746807.992316] [1721397] 1000 1721397 33547921 11490051 99647488 0 0 claude
[6746807.992318] [1732461] 0 1732461 3860 1013 73728 0 0 sshd
[6746807.992320] [1732462] 109 1732462 3860 681 77824 0 0 sshd
[6746807.992322] [1734802] 0 1734802 3860 1108 73728 0 0 sshd
[6746807.992324] [1734813] 1000 1734813 698 248 49152 0 0 sleep
[6746807.992326] [1734836] 109 1734836 3860 695 73728 0 0 sshd
[6746807.992328] [1735021] 1000 1735021 723 221 49152 0 0 sh
[6746807.992330] [1735022] 1000 1735022 1091 495 53248 0 0 cpuUsage.sh
[6746807.992332] [1735025] 1000 1735025 1548 181 49152 0 0 sleep
[6746807.992334] [1735030] 1000 1735030 698 244 40960 0 0 sleep
[6746807.992336] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=tailscaled.service,mems_allowed=0,global_oom,task_memcg=/user.slice/user-1000.slice/session-c88.scope,task=claude,pid=1721397,uid=1000
[6746807.992366] Out of memory: Killed process 1721397 (claude) total-vm:134191684kB, anon-rss:45957644kB, file-rss:2560kB, shmem-rss:0kB, UID:1000 pgtables:97312kB oom_score_adj:0
● This is an Out of Memory (OOM) kill event. The system killed the claude process (PID 1721397) because it was consuming excessive memory.
Key details:
- Process killed: claude (PID 1721397, UID 1000)
- Memory usage: 134,191,684 kB (128 GB) virtual memory, 45,957,644 kB (44 GB) resident memory
- Reason: System ran out of memory (global OOM)
- Time: 6746807.992366 seconds since boot
The process list shows claude was by far the largest memory consumer. Other processes like node and dockerd were using much less memory. The OOM killer selected claude as the best candidate to free up memory and keep
the system operational.
Metadata
Metadata
Assignees
Labels
area:corebugSomething isn't workingSomething isn't workinghas reproHas detailed reproduction stepsHas detailed reproduction stepsmemoryperf:memoryplatform:linuxIssue specifically occurs on LinuxIssue specifically occurs on Linux
Activity
mark-99 commentedon Jul 16, 2025
I did install this checkpoint util recently, and I was accessing some large log files, and I see it's inserted a hook, so that's a possibility.
claude-code-checkpoint
Here's my full
settings.jsonmark-99 commentedon Jul 16, 2025
So downgraded to 1.0.52 and I didn't have any problems.
I uninstalled the checkpoint util (hook was removed from the settings.json), upgraded to 1.0.53 and it OOMed again:
[6752564.571263] Out of memory: Killed process 2028082 (claude) total-vm:133718668kB, anon-rss:44305312kB, file-rss:2444kB, shmem-rss:0kB, UID:1000 pgtables:94080kB oom_score_adj:0mark-99 commentedon Jul 16, 2025
Happening on 1.0.52 also :/
mark-99 commentedon Jul 18, 2025
Reverted further to 1.0.51 and no issues in the last 24 hours.
bcherny commentedon Jul 27, 2025
Actively working on squashing memory leaks!
github-actions commentedon Aug 5, 2025
This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.