May I ask for some help about mongodb memory usage? Thanks in advance!
I have a production mongodb instance running in the docker compose. The mongodb status is as follows,
> db.serverStatus().tcmalloc.tcmalloc.formattedString
------------------------------------------------
MALLOC: 612502384 ( 584.1 MiB) Bytes in use by application
MALLOC: + 936054784 ( 892.7 MiB) Bytes in page heap freelist
MALLOC: + 49701472 ( 47.4 MiB) Bytes in central cache freelist
MALLOC: + 5157248 ( 4.9 MiB) Bytes in transfer cache freelist
MALLOC: + 11567280 ( 11.0 MiB) Bytes in thread cache freelists
MALLOC: + 9568256 ( 9.1 MiB) Bytes in malloc metadata
MALLOC: ------------
MALLOC: = 1624551424 ( 1549.3 MiB) Actual memory used (physical + swap)
MALLOC: + 311873536 ( 297.4 MiB) Bytes released to OS (aka unmapped)
MALLOC: ------------
MALLOC: = 1936424960 ( 1846.7 MiB) Virtual address space used
MALLOC:
MALLOC: 29704 Spans in use
MALLOC: 274 Thread heaps in use
MALLOC: 4096 Tcmalloc page size
------------------------------------------------
Call ReleaseFreeMemory() to release freelist memory to the OS (via madvise()).
Bytes released to the OS take up virtual address space but no physical memory.
IIUC, it uses 1846.7 MiB memory. But when I check the docker status,
# docker stats --no-stream
CONTAINER ID NAME CPU % MEM USAGE / LIMIT MEM % NET I/O BLOCK I/O PIDS
56b9bd33c533 slip-backend-web-1 0.00% 12.32MiB / 7.447GiB 0.16% 142MB / 183MB 0B / 0B 3
3f2c6a129d91 slip-backend-backend-1 0.08% 161.1MiB / 7.447GiB 2.11% 239MB / 210MB 0B / 0B 5
6afbcbf0dddc slip-backend-mongo-1 0.61% 4.212GiB / 7.447GiB 56.57% 45.9GB / 202GB 1.88GB / 798GB 650
I found it consums 4.212GiB memory.
Could you give me a hand to figure out why there is a memory difference and how to avoid it?
From the following command, we can see that the mongodb container has been running for over 2 months, not sure if it's related.
# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
56b9bd33c533 nginx "/docker-entrypoint.…" 2 hours ago Up 2 hours 0.0.0.0:80->80/tcp, :::80->80/tcp, 0.0.0.0:443->443/tcp, :::443->443/tcp, 0.0.0.0:8123->8133/tcp, :::8123->8133/tcp slip-backend-web-1
3f2c6a129d91 slip-backend-backend "python3 server.py" 2 hours ago Up 2 hours slip-backend-backend-1
6afbcbf0dddc mongo "docker-entrypoint.s…" 8 weeks ago Up 8 weeks 0.0.0.0:321->321/tcp, :::321->321/tcp, 27017/tcp
I've found a temporary workaround by restarting the MongoDB container using the command
docker compose restart mongo
and then,
# docker stats --no-stream
CONTAINER ID NAME CPU % MEM USAGE / LIMIT MEM % NET I/O BLOCK I/O PIDS
56b9bd33c533 slip-backend-web-1 0.00% 14.61MiB / 7.447GiB 0.19% 191MB / 247MB 0B / 0B 3
3f2c6a129d91 slip-backend-backend-1 0.12% 185MiB / 7.447GiB 2.43% 318MB / 283MB 0B / 0B 5
6afbcbf0dddc slip-backend-mongo-1 0.72% 243.3MiB / 7.447GiB 3.19% 8.65MB / 9.71MB 23.9MB / 38.6MB 38
While I haven't yet identified the root cause, it appears that over time, the memory used by the MongoDB container isn't being properly released, leading to increasing memory consumption. It seems like a memory leak or a problem with memory management within the container.
Regardless, I believe this is a valuable issue to document, and I hope this workaround helps others who encounter similar problems.