0

I have two servers (Oracle Linux 7, x86_64, ) with "missed" RAM.

uname -r returns 4.14.35-1902.10.2.1.el7uek.x86_64

I've read where is my memory in linux? and already checked memory by the script (mema.py)

The script shows the following result:

Linux memory analyser v0.1

Usable RAM total:        15.317 GB
Used RAM:                15.066 GB
Used by bufers+cache:    2.763 GB
Used by programs:      6.308 GB (23.80%)
Shared or unknown:     5.994 GB (39.13%) (WTF?)

By the time system uptime was 126 days and all went well (no OOM killer, no lags, just customer's questions). It seems like cached memory isn't visible (just like with ZFS, but the servers use XFS).

lsmod doesn't show any module with name like zfs.

I've tried to set 2 and 3 to vm.drop_caches via systctl - no any effect for Shared or unknown (but Used by bufers+cache actually decreased).

After reboot Shared or unknown decreased to 1.153 GB (7.53%) (Good) but it grows gradually.

How can I check that the issue is like ZFS' issue (or not)?

I will appreciate for any help.

1
  • You can get detailed reports on shared memory with ipcs -m and other options shown in the man page. Some systems may have smem instead. Feb 4 at 16:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.