memory logs
 

Hi,

I manually created a process to occupy the memory and when i tried to monitor the logs when physical memory is full, i was not able to see any "alert" or "warning" or "oom" or "out of memory" in logs.

Kindly let me know where can i check the logs when the physical memory is full ???

[root@localhost sa]# uname -a
Linux localhost.localdomain 3.10.0-123.el7.x86_64 #1 SMP Mon May 5 11:16:57 EDT 2014 x86_64 x86_64 x86_64 GNU/Linux

root@localhost /]# cat: write error: No space left on device
ls -ltr
total 17348580

[root@localhost /]# df -TH
Filesystem Type Size Used Avail Use% Mounted on
/dev/sda3 xfs 20G 20G 21k 100% /
devtmpfs devtmpfs 953M 0 953M 0% /dev
tmpfs tmpfs 958M 0 958M 0% /dev/shm
tmpfs tmpfs 958M 9.1M 949M 1% /run
tmpfs tmpfs 958M 0 958M 0% /sys/fs/cgroup
/dev/sda1 xfs 312M 89M 223M 29% /boot

no luck on using below commands...

[root@localhost /]# egrep -ir "alert|warning|oom|out of memory" /var/log/*
[root@localhost /]# journalctl -p warning

you can check, if the swap is in use the main memory usually is full. Otherwise you ought to read about memory handling on linux, it is not that simple at all. For example you can try to allocate more memory and even more memory (more than what you have).

Hi,

I understand your point.

Here I am saying about permanent memory (hdd) and not about temporary memory(RAM, SWAP)

When my HDD is full , the logs should have the records of every transaction, right ?

when the root filesystem is full probably you cannot write log files...

Why?

Show us your code and we'll show you ours? ;)

test_fill /boot partition, instead of root fs (where your /var/log is)
 

[wild-guess] Try `dmesg|tail`. Per `man dmesg`:
When fs is full, you cannot write to it, so you can't read the info that you canNOT write.
p.s. disk is NOT usually called "physical memory"; RAM is called memory; disk is called disk. See wiki. (Was your *goal* to study 'full' errors? Have /var/log on *separate* partition!)

reboot.

Done.

clean out /home as it appears to be part of the 20 G / (sda5 ) partition.
should show you something.

should show you "more" of something.

@OP: change your `cat ... >file&` *TO* >/boot/file&
Pick a *test file name*. like zzz or MYfile, that doesNOT overwrite/damage anything already there!

Then, try your egrep and journalctl commands. Let us know if that helps.

p.s. your /home may be empty, if you didn't put anything there; mine (tempVBoxes) often is;
`reboot` may loose/erase/destroy your "kernel ring buffer", but I'm not sure.
(I think SunOS4.x used to save it across reboots, for logging, but I haven't taken the time to try/study this on Linux.)

Also, please Edit in Post#1 here the Title of this Thread to be more accurately descriptive, like:
Want to see/view system log error messages when a filesystem is full/filled (learning)

This will help others find this solution in the future, esp. via. "Similar Threads" at bottom of page. Also, use CODE tags. ('bbcode' formatting)

probably you can remove: /var/cache/apt/archives/*
although it is not a nice solution
by the way, what will du -sh /var/* respond?