Pleskrestore is broken, how to fix?

[ralphschindler]

I created a backup file on a CentOS4/Plesk8.2 box and am trying to restore it on a Redhat5/Plesk8.2 box.

I get the following error:
[root@plesk /]# /usr/local/psa/bin/pleskrestore --restore /home/admin/pleskbackup82.file -map /home/admin/backup82.map -level server -license -verbose
WARNING: Backup file has wrong signature for this server

WARNING! Restore with "-level server" switch brings server to exactly the same
state, as was saved in backup file. All objects, not presented in the backup
file (clients or domains) will be deleted. Proceed? [yes/no]: yes
Uncaught std::exception: UnixClientSocket::UnixClientSocket, connect: No such file or directory
Traceback (most recent call last):
  File "/usr/local/psa/admin/share/supervisor/processor.py", line 116, in doActivityRunner
    newState = self.doActivity()
  File "/usr/local/psa/admin/share/supervisor/processor.py", line 817, in doActivity
    cmd.spawn(os.environ.has_key("PLESKRESTORE_MODE"))
  File "/usr/local/psa/admin/lib/python/subproc.py", line 238, in spawn
    proc.run()
  File "/usr/local/psa/admin/lib/python/subproc.py", line 193, in run
    BaseSubprocess.run(self)
  File "/usr/local/psa/admin/lib/python/subproc.py", line 177, in run
    self.wait()
  File "/usr/local/psa/admin/lib/python/subproc.py", line 197, in wait
    BaseSubprocess.wait(self)
  File "/usr/local/psa/admin/lib/python/subproc.py", line 187, in wait
    raise NonzeroExitException(self, os.WEXITSTATUS(status))
NonzeroExitException:  was finished with exit code 1


[root@plesk /]# cat /usr/local/psa/version 
8.2.0 RedHat el5 82070807.17

What gives?

 

[shemso]

I have the same problem with "Backup file has wrong signature for this server" ...
did you figure how to fix this situation?

 

[anb_t]

Originally posted by ralphschindler
I created a backup file on a CentOS4/Plesk8.2 box and am trying to restore it on a Redhat5/Plesk8.2 box.

I get the following error:
[root@plesk /]# /usr/local/psa/bin/pleskrestore --restore /home/admin/pleskbackup82.file -map /home/admin/backup82.map -level server -license -verbose
WARNING: Backup file has wrong signature for this server

WARNING! Restore with "-level server" switch brings server to exactly the same
state, as was saved in backup file. All objects, not presented in the backup
file (clients or domains) will be deleted. Proceed? [yes/no]: yes
Uncaught std::exception: UnixClientSocket::UnixClientSocket, connect: No such file or directory
Traceback (most recent call last):
  File "/usr/local/psa/admin/share/supervisor/processor.py", line 116, in doActivityRunner
    newState = self.doActivity()
  File "/usr/local/psa/admin/share/supervisor/processor.py", line 817, in doActivity
    cmd.spawn(os.environ.has_key("PLESKRESTORE_MODE"))
  File "/usr/local/psa/admin/lib/python/subproc.py", line 238, in spawn
    proc.run()
  File "/usr/local/psa/admin/lib/python/subproc.py", line 193, in run
    BaseSubprocess.run(self)
  File "/usr/local/psa/admin/lib/python/subproc.py", line 177, in run
    self.wait()
  File "/usr/local/psa/admin/lib/python/subproc.py", line 197, in wait
    BaseSubprocess.wait(self)
  File "/usr/local/psa/admin/lib/python/subproc.py", line 187, in wait
    raise NonzeroExitException(self, os.WEXITSTATUS(status))
NonzeroExitException:  was finished with exit code 1


[root@plesk /]# cat /usr/local/psa/version 
8.2.0 RedHat el5 82070807.17

What gives?

What 'pleskrestore --info' says about the backup file?

/usr/local/psa/bin/pleskrestore --info /home/admin/pleskbackup82.file

./anb

 

[plugged]

i had the same problem today. i had secured the /tmp with noexec etc... but therefor the unixsocket didn't worked anymore... unsecuring the /tmp and it worked again....

 

[svaughan]

FYI, I had the same problem. When you first run the pleskrestore script it creates a folder called 'restore' in the current directory, you need to give that folder 777 permissions, or something close to that. So make the restore folder prior to running the restore script and give it the permissions. Or you might just have to chown it for one of the psa users, either way it should fix it.

You can always run the script and chmod it afterwards and then re-run pleskrestore

 

[blank_page]

HI,

I fixed this with yum update perl-DBI

the backup log should point you to the right directions though.

 

[GLehnhoff@]

Solution

This helped me:

# known problem to Plesk ... find file usr/local/psa/PMM/plesk.dtd.
# Search for ...
#
# <!ATTLIST pinfo
# name (company|phone|fax
# |address|city|state
# |zip|country|locale|email
# |name) #REQUIRED
#
# ... and replace line ...
#
# |name) #REQUIRED
#
# ... with ...
#
# |name|info_not_required) #REQUIRED

 

[mousehouse]

After having a domain that refused to be transferred using the migration manager (PMM) because a perl job would hang on the source server I reverted back to using pleskbackup and pleskrestore. This was no easy route, as the source system was on Plesk 8.2.1. with a b0rked install that would not upgrade to any older Plesk version (thanks to so-called Plesk experts doing the job). The destination system Plesk 8.6 ;-)

Pleskrestore would not do the job, and in the end I was able to restore by doing two thing,

1) The excellent trick described above.

2) Move the backupfile to "/root" and doing the pleskrestore from there.

This did the trick and got my last 8.2.1 domain migrated!

Thanks to all for this post, saved my day.