Rm file will cause freenas crash restart
Root@freenas:/mnt/pool1/ghomepcbak # date
Mon Aug 20 21:54:43 CST 2018
Root@freenas:/mnt/pool1/ghomepcbak # rm -rf *
Socket error Event: 32 Error: 10053.
Connection closing...Socket close.
Connection closed by foreign host.
Disconnected from remote host(FreeNas@gen8) at 21:56:11.
Type `help' to learn how to use Xshell prompt.
Will cause freenas to die and need to be restarted. Retrying three times in a row will trigger this issue each time.
After the problem occurs, ssh, http can not be connected, the display still displays the console settings and other content, no error message, the local keyboard click does not have any corresponding. Waiting for a while will display the attachment information, which should be followed by an automatic restart.
I will execute the zpool scrub to check if there is an error in the pool, but it shouldn't be a problem, because it was just executed a few days ago and no errors were checked.
Do you still need to provide additional content?
I will try to delete ghomepcbakdataset directly after I don't need to provide other content.
#3 Updated by gamexg gamexg about 2 years ago
See #40928, I am also over 80% of the capacity here, trying to free up space by attempting to clear backups that are no longer needed.
I don't know if I should keep the current fault environment to provide more information about the crash or to free up space and try to reproduce the crash?
#7 Updated by Alexander Motin about 2 years ago
- Status changed from Unscreened to Closed
- Target version changed from Backlog to N/A
- Reason for Closing set to Duplicate Issue
- Needs QA changed from Yes to No
- Needs Doc changed from Yes to No
- Needs Merging changed from Yes to No
It is a full duplicate of bug #41910. I am investigating it already. The system seem to be too "busy" to respond. Out of 4 CPU cores one looks like working on deletion (though I am not completely sure why it doesn't finish), while all the rest are spinning waiting for that, making system not responsive until watchdog timer fire.
You may try deleting one file at a time. I suppose it should work until better fix is available.