#1600 Issue closed
: SSHFS ReaR Hangs¶
Labels: support / question
, not ReaR / invalid
John-Leone opened issue at 2017-11-27 19:50:¶
Relax-and-Recover (ReaR) Issue Template¶
Fill in the following items before submitting a new issue
(quick response is not guaranteed with free support):
[root@xhcsm5a rear]# rear -V
Relax-and-Recover 2.00 / Git
Red Hat Enterprise Linux Server release 7.3 (Maipo)
OUTPUT=ISO
BACKUP=NETFS
OUTPUT_URL=sshfs://root@xxx.xxx.xx.xxx/dbar/
BACKUP_URL=sshfs://root@xxx.xxx.xx.xxx/dbar/
KEEP_OLD_OUTPUT_COPY=1
BACKUP_PROG_EXCLUDE=("${BACKUP_PROG_EXCLUDE[@]}" '/u01'
'/var/cache/yum')
using legacy BIOS boot?
We use ReaR with sshfs option to backup Linux physical/virtual servers. This works very well but each week we have approximately 30 backups fail because sshfs hangs.
The current work around is to create the ReaR files local on the server then copy them to the backup server.
Here is the ls output from /tmp:
[root@xhcsm5a rear.tdHIIt4LVxXrsKA]# pwd
/tmp/rear.tdHIIt4LVxXrsKA
[root@xhcsm5a rear.tdHIIt4LVxXrsKA]# ls -lhart
ls: cannot access outputfs: Transport endpoint is not connected
total 0
??????????? ? ? ? ? ? outputfs
drwxr-xr-x 2 root root 6 Nov 27 14:16 tmp
drwxr-xr-x 3 root root 16 Nov 27 14:16 rootfs
drwx------ 5 root root 44 Nov 27 14:16 .
drwxrwxrwt. 4 root root 50 Nov 27 14:25 ..
In most cases a server reboot will fix the problem but I'm wondering if there was anyway to avoid rebooting the server to clear this issue?
Thanks,
John
rear-xhcsm5a.log
gozora commented at 2017-11-27 20:09:¶
Hello @John-Leone,
This don't look like ReaR issue to me.
You can confirm directly with RedHat, but AFAIK sshfs is not supported
on RHEL...
V.
John-Leone commented at 2017-11-27 20:16:¶
Hi @gozora
That's correct sshfs is not supported by RedHat, I opened a case with
them and they don't support sshfs.
I'm looking for a workaround other than rebooting the server.
gozora commented at 2017-11-27 20:25:¶
Lot of things can be wrong here and I guess RedHat does not support
sshfs for a reason ...
I personally did not run into any kind of hang situations with FUSE
filesystems, but try just umount stale share:
fusermount -u /tmp/rear.tdHIIt4LVxXrsKA/outputfs
V.
John-Leone commented at 2017-11-27 20:44:¶
sshfs works fine most of the time.
I tried fusermount -u then tried running a new backup with the same
result.
gozora commented at 2017-11-27 20:48:¶
Most of the time ... :-).
Did you try rmmod fuse && modprobe fuse
?
V.
John-Leone commented at 2017-11-27 21:01:¶
That's a good idea, but same result.
gdha commented at 2017-11-28 07:48:¶
@John-Leone Please add ServerAliveInterval 15
to your .ssh/config
file to avoid such issues in the future
jsmeix commented at 2017-12-01 13:32:¶
I close it because it is not an issue in ReaR.
[Export of Github issue for rear/rear.]