Issue Restoring backup stuck

[Lanta]

yesterday i tried restoring a backup and now the process seems to be stuck.
Plesk is showing this message for the last 2 days when in the backup manager:
Restoring Feb 13, 2017 02:30 PM
Tried restarting the server didnt work

OS‪CentOS 6.8 (Final)‬ProductPlesk Onyx
Version 17.0.17 Update #10, last updated on Dec 11, 2016 09:09 PM

 

[Lanta]

Updated to

Plesk Onyx
Version 17.0.17 Update #16, last updated on Feb 14, 2017 12:55 PM

issue still there

 

[Bitpalast]

Please check the restore log files of the hanging restore for additional information what went wrong in /usr/local/psa/PMM/logs/restore-*

Further actions depend on the situation.

 

[Lanta]

Hi Peter,

Thanks for your reply.

I have looked at the log files and i can't see what is going wrong. To me it seems like the process just stopped.

these are the last lines of deployer.log

[2017-02-13 14:30:00.207|26384] INFO: Repository '/var/lib/psa/dumps/': Validate backup (common phase) clients/opzeker/backup_info_1702080947.xml
[2017-02-13 14:30:00.247|26384] INFO: Repository '/var/lib/psa/dumps/': Put info to clients/opzeker/.discovered/backup_info_1702080947/status_OK
[2017-02-13 14:30:00.259|26384] INFO: pmm-ras finished. Exit code: 0
[2017-02-13 14:30:00.304|26462] INFO: pmm-ras started : /usr/local/psa/admin/bin/pmm-ras --get-dump-info --dump-storage=/var/lib/psa/dumps/ --dump-file-specification=backup_info_1702080947.xml --session-path=/var/log/plesk/PMM
[2017-02-13 14:30:00.305|26462] INFO: Repository '/var/lib/psa/dumps/': Initializing...
[2017-02-13 14:30:00.306|26462] INFO: Repository '/var/lib/psa/dumps/': Initialized
[2017-02-13 14:30:00.306|26462] INFO: Repository '/var/lib/psa/dumps/': Get backup info from backup_info_1702080947.xml
[2017-02-13 14:30:00.306|26462] INFO: Repository '/var/lib/psa/dumps/': Get initial backup info for backup_info_1702080947.xml
[2017-02-13 14:30:00.306|26462] INFO: Repository '/var/lib/psa/dumps/': Get info from .discovered/backup_info_1702080947/props
[2017-02-13 14:30:00.306|26462] INFO: Read properties from xml backup_info_1702080947.xml
[2017-02-13 14:30:00.308|26462] INFO: Repository '/var/lib/psa/dumps/': Validate backup backup_info_1702080947.xml
[2017-02-13 14:30:00.308|26462] INFO: Repository '/var/lib/psa/dumps/': Get info from .discovered/backup_info_1702080947/props
[2017-02-13 14:30:00.308|26462] INFO: Repository '/var/lib/psa/dumps/': Get info from .discovered/backup_info_1702080947/status_OK
[2017-02-13 14:30:00.309|26462] INFO: Repository '/var/lib/psa/dumps/': Get info from .discovered/backup_info_1702080947/dumpresult_SUCCESS
[2017-02-13 14:30:00.309|26462] INFO: pmm-ras finished. Exit code: 0
[2017-02-13 14:30:00.359|26463] INFO: pmm-ras started : /usr/local/psa/admin/bin/pmm-ras --get-child-dumps --dump-storage=/var/lib/psa/dumps/ --dump-specification=backup_info_1702080947.xml --session-path=/var/log/plesk/PMM
[2017-02-13 14:30:00.360|26463] INFO: Repository '/var/lib/psa/dumps/': Initializing...
[2017-02-13 14:30:00.360|26463] INFO: Repository '/var/lib/psa/dumps/': Initialized
[2017-02-13 14:30:00.360|26463] INFO: Repository '/var/lib/psa/dumps/': Get all child backups for backup_info_1702080947.xml
[2017-02-13 14:30:00.411|26463] INFO: pmm-ras finished. Exit code: 0

there was also a pmmcli.log file in the directory (see attachment)

on line 51 i see

[2017-02-13 14:30:02.368|26382] INFO: Subprocess raised ExecuteException: Subprocess <subprocess[26467] '/usr/local/psa/bin/sw-engine-pleskrun -c /usr/local/psa/admin/conf/php.ini /usr/local/psa/admin/plib/backup/Conflicts/Runner.php --fix-guids --restore-specification=/usr/local/psa/PMM/rsessions/20170213143000316/dump/dump_index.xml --session-path=/usr/local/psa/PMM/rsessions/20170213143000316'> was finished with exit code 254

 

[Bitpalast]

If I only knew what the "GUID fixer" is doing and what "exit code 254" indicates for that ...

Actually, it looks as if the restore was finished, because final exit codes are zero. Have the files and settings been restored? In that case it might be sufficient to remove the job from longtasks table in the database and maybe kill running PMM tasks. At least this should clear the running "hourglass" message.

1) Backup psa database (KB #213904125)

2) # plesk db
> select * from longtasks;
See if the pmm task is in there, should be the only one or one out of few. If things look generally "ok", proceed
> truncate longtasks;
> truncate longtaskparams;
to remove all entries from these tables.

3) # ps aux | grep pmm
to find pmm processes, then kill them by pid
# kill -p PID

4) Retry restore, if needed.

Maybe you need to open a support ticket if noone else on here can point into a better direction.

 

[Lanta]

Thank you Peter!

Unfortunately it didn't resolve the issue

mysql> select * from longtasks
mysql> truncate longtasks;
Query OK, 0 rows affected (0.01 sec)
mysql> truncate longtaskparams;
Query OK, 0 rows affected (0.01 sec)

seems there was nothing in the table.
Also the files where never restored.
I will try to restore the domain again but now i will download the backup files from the ftp to the local backup repository first. (will take ages because the backup files 18GB are stored on a ftp server with a slow connection 30mbit and plesk stores all the domains in 1 single backup file...)

 

[Lanta]

this time the restore was succesfull. But i still got the message saying "Restoring Feb 13, 2017 02:30 PM". Is there a way to remove this message?