1. Please take a little time for this simple survey! Thank you for participating!
    Dismiss Notice
  2. Dear Pleskians, please read this carefully! New attachments and other rules Thank you!
    Dismiss Notice
  3. Dear Pleskians, I really hope that you will share your opinion in this Special topic for chatter about Plesk in the Clouds. Thank you!
    Dismiss Notice

Migration Manager trouble

Discussion in 'Plesk for Linux - 8.x and Older' started by wizbones, Apr 18, 2006.

  1. wizbones

    wizbones Guest

    0
     
    When I try to migrate from my Raq4 I get the following error:

    As root:

    Error during dump retrieval


    Error during dump file retrieving.
    Remote agent failed with the following stderr messages: 10:53:45 INFO FileFinder: psql resolved to /usr/bin/psql
    10:53:45 INFO FileFinder: pg_dump resolved to /usr/bin/pg_dump
    10:53:45 INFO New connection: postgresql shell connection. db cobalt, user admin, host localhost
    10:53:45 INFO ------------------------------------------------------------
    10:53:45 INFO FILE storage initialized.
    10:53:45 INFO Base directory: /home/migration/supervisor.4588/
    10:53:45 INFO Gzip bundles: yes
    10:53:45 INFO Bundle split size: 1073741824
    10:53:45 INFO ------------------------------------------------------------
    10:53:45 TRACE SQL: select name,fqdn,hostname,domain from vsite
    psql: PQconnectPoll() -- connect() failed: Connection refused
    Is the postmaster running (with -i) at 'localhost'
    and accepting connections on TCP/IP port '5432'?
    10:53:45 TRACE SQL: SELECT v.name, MIN(u.name) FROM vsite v, users u WHERE u.vsite=v.name AND (u.admin='t') GROUP by v.name
    psql: PQconnectPoll() -- connect() failed: Connection refused
    Is the postmaster running (with -i) at 'localhost'
    and accepting connections on TCP/IP port '5432'?
    10:53:45 TRACE SQL: SELECT name FROM vsite
    psql: PQconnectPoll() -- connect() failed: Connection refused
    Is the postmaster running (with -i) at 'localhost'
    and accepting connections on TCP/IP port '5432'?
    10:53:45 TRACE SQL: SELECT COUNT(*) FROM users WHERE vsite = 'dbvsites' AND users.admin='t'
    psql: PQconnectPoll() -- connect() failed: Connection refused
    Is the postmaster running (with -i) at 'localhost'
    and accepting connections on TCP/IP port '5432'?
    Can't use an undefined value as an ARRAY reference at /home/migration/supervisor.4588/RaQ4.pl line 336.

    As admin:

    Cannot send scout to the remote host


    Failed to execute scout.
    The return code is 127
    The output on STDERR is
    sh: sudo: command not found
    sh: sudo: command not found

    Has anyone successfully migrated from a Raq? I would really like to get this working so I can retire that old blue box. Any suggestions would be greatly appreciated.
     
  2. dm__@

    dm__@ Guest

    0
     
    Read the logs:

    Is the postmaster running (with -i) at 'localhost'
    and accepting connections on TCP/IP port '5432'?
     
  3. wizbones

    wizbones Guest

    0
     
    I've checked everywhere I can think and nope not running with -i and is excepting connections on that port:

    Here is the postmaster.opts file:

    /usr/bin/postmaster
    -p 5432
    -D /var/lib/pgsql/data
    -B 64
    -b /usr/bin/postgres
    -N 32

    As far as I've read even if I were to change this file as soon as it's restarted it will revert to this. I've seen other ones with the -i option, which I do not. What am I missing?

    What log should I be checking. Postgresql in /var/log is empty, so is the other two compressed archives of the log.
     
  4. dbase

    dbase Guest

    0
     
    I noticed that postmaster.opts get overwritten when a /etc/rc.d/init.d/postgresql restart is issued.

    Just a quick workaround (probely not the best way):
    Edit /etc/rc.d/init.d/postgresql and add the -i flag after the start command on line 62:

    Code:
    [root@raq2 init.d]# diff postgresql postgresql.bup
    62c62
    <       su postgres -c "nohup /usr/bin/postmaster -i >> /var/log/postgresql 2>&1 &"
    ---
    >       su postgres -c "nohup /usr/bin/postmaster >> /var/log/postgresql 2>&1 &"
    [root@raq2 init.d]#
    
    restart the service and check with telnet:
    Code:
     
    [root@raq2 init.d]# /etc/rc.d/init.d/postgresql restart
    Shutting down PostgreSQL: ok
    Setting up PostgreSQL: ok [20748]
    [root@raq2 init.d]# telnet 0 5432
    Trying 0.0.0.0...
    Connected to 0.
    Escape character is '^]'.
    ^]
    telnet> close
    Connection closed.
    [root@raq2 init.d]#
    
    I hope this helps.

    Best regards,
    Bas Hendriks
     
  5. cschnee

    cschnee Guest

    0
     
    Thanks for the tips. But even with postgres enabled for tcp with -i as described and a allow line in pg_hba.conf I still get

    "Failed validation of the document with the external DTD /opt/psa/PMM/agents/RaQ4/control/RaQ4.dtd. The errors are: Element RaQ4dump does not carry attribute agent-name"

    So I guess concerning Cobalts the new Migration Tool of 8.0x is not working at all.

    Anybody managed to migrate RaQs with Plesk 8 ?
     
  6. wizbones

    wizbones Guest

    0
     
    Still no luck here at all. Batting Zero
     
  7. cschnee

    cschnee Guest

    0
     
    Im going back to 7.54 where it worked, but the uninstallation of all plesk components on a debian system turns out to be a major PITA...
     
  8. wizbones

    wizbones Guest

    0
     
    Ok, using the -i for postgresql and editing the pg_hba.conf file I finally got the same error as cschnee

    "Failed validation of the document with the external DTD /opt/psa/PMM/agents/RaQ4/control/RaQ4.dtd. The errors are: Element RaQ4dump does not carry attribute agent-name"

    Really would like to get this working. We're getting there one step at a time.

    Anyone get past this point?
     
  9. dm__@

    dm__@ Guest

    0
     
    That's the time to ask support for help: as this looks like Plesk bug they should not charge you. But it's always advisable to install all available updates first.
     
  10. wizbones

    wizbones Guest

    0
     
    Took it one step further, on the RaQ4 the dump.xml was created and all the domains/users are listed. So I tried copying the xml file to the plesk box and running a local import to see what would happen.

    The error: 'agent-name' attribute is not found in dump file

    So looks like that's the trouble will send this on to support and see what they say.

    Thanks dm__ for your help.
     
  11. wizbones

    wizbones Guest

    0
     
    Submitted my trouble ticket:

    Was sent a file, replaced it, didn't work. Was sent another file and my antivirus program kept eating it. Guess it was hungry.

    Anyhow, Plesk wanted access to my server to install an updated migration package for RaQ - well I am currently in the process of moving two domains from my RaQ to my Plesk.

    They seem to be working but also kinda sorta seem stuck on moving the web portion.

    Will be patient and let this run until it works or errors. Will post back results when completed either way.

    Here's to success - cheers and laters.
     
  12. wizbones

    wizbones Guest

    0
     
    After several attempts which failed I finally decided to restart the RaQ server. I kept getting stuck at transferring the web portion of the migration.

    After restart - bam - it's a working. Thanks Plesk support.

    What did they do exactly? Changed my migration installation somehow. Not sure exactly - but if you cannot migrate from a RaQ I highly recommend sending in a trouble ticket now, and give them access - they can fix it rather quickly.

    Thanks again support guys. I have started to move and getting ready to retire that tired old RaQ system.
     
Loading...