12-27-2022 05:06 AM
Good morning. The /log directory on my appliance is full. It's looking like much of the space is used by old core files in /log/crash. Can those be deleted?
12-27-2022 05:18 AM
yes these files can be deleted
for more information, check this : https://www.veritas.com/support/en_US/article.100046784
12-27-2022 05:21 AM
Thank you for the quick reply. I removed some of the files as a test and while it shows used space has dropped it's still reporting 100% full.
/dev/mapper/system-log 182G 179G 0 100% /log
Is there a service I need to restart for it to recognize the space again?
12-27-2022 05:40 AM
indeed that is strange..
it shows 179G or 182G is used, so at least 3G should be free..
try to reboot the server..
also, check if you still have dump or crash files created in the folder..
12-27-2022 02:18 PM
You need to remember that in Linux, a percentage of the file system sapce is reserved so although you have 3GB free space, this still registers as 100% full (the perentage can vary and is typically 10%).
If the reboot hasn't done anything (which I suspect it wont), use the du command in the log directory to identify the directories consuming all the space. The idea is starting from the /log folder, run the command "du -sh *" - this will produce a listing of all the folders/files in the /log folder with the total size in human readable format. Then drill down each of the large directories to identify where all the space is being consumed.
There is little required information contained in the log directory, so removing old log files is fine. I would also remove ALL core files (although identifying what is generating them first would be a good idea as there may be an EEB to address it).
12-28-2022 12:20 AM
Also run NBU command to delete old logs :
vxlogmgr --del -p 51216 -b '12/28/2022 12:00:00 PM'
vxlogmgr does not delete core files.
01-04-2023 12:20 PM
Delete all the core files and remove all the log files older than 2 days. You can also just delete everything in the log directory.