ISPconfig migration tool?

Discussion in 'Plugins/Modules/Addons' started by jboud, Jan 17, 2017.

Page 3 of 5
  1. reason8

    reason8 Member

    After disabling SSL, this is the new messge I get with wget:

    wget http://IPADDR:8080/remote/

    --2017-07-12 20:25:49-- http://IPADDR:8080/remote/
    Connecting to IPADDR:8080... connected.
    HTTP request sent, awaiting response... 400 Bad Request
    2017-07-12 20:25:49 ERROR 400: Bad Request.
     
    reason8, Jul 13, 2017
    #41
  2. till

    till Super Moderator Staff Member ISPConfig Developer

    The migration tool has no --no-check-certificate option, that's an option of the wget command. You said that you disabled SSL to circumvent that but SSL seems still to be on. Please edit the ispconfig vhost file on the new server and disable SSL there. Then try if you can connect with wget and http to http://IPADDR:8080/remote/
    When this is working, then try to run the migration tool again and use the http://IPADDR:8080/remote/ as the remoting API address.

    Just a side note, the migration tool support can be contacted here directly: http://www.ispconfig.org/get-support/?type=migration
     
    till, Jul 13, 2017
    #42
  3. reason8

    reason8 Member

    I disabled SSL again (I turned it on after the last attempt)

    wget http://IPADDR:8080/remote/
    --2017-07-13 07:23:21-- http://IPADDR:8080/remote/
    Connecting to IPADDR:8080... connected.
    HTTP request sent, awaiting response... 500 Internal Server Error
    2017-07-13 07:23:21 ERROR 500: Internal Server Error.

    Still cant connect from the Source to the Target.

    migrate.log shows:
    2017-07-13 14:23:45 - [INFO] Migration tool version 1.6.10 started.
    2017-07-13 14:23:46 - [INFO] /root/migration/legacy/conf/importer.conf.php does not seem to contain a valid configuration.
    2017-07-13 14:25:27 - [ERROR]
     
    reason8, Jul 13, 2017
    #43
  4. till

    till Super Moderator Staff Member ISPConfig Developer

    ok, that's good, as it means that there is a problem on the new server. So we are one step further :) Maybe the php soap package is missing, try to install it. You can find the detailed reason in the apache error.log on the new server.
     
    till, Jul 13, 2017
    #44
  5. reason8

    reason8 Member

    I followed the Perfect Server installation for Ubuntu 17. Not sure why it would not be installed.

    Here is the message for the apache2 error log:
    [Thu Jul 13 07:28:48.270741 2017] [fcgid:warn] [pid 6777] [client SOURCE_IPADDR:53430] mod_fcgid: stderr: thrown in /usr/local/ispconfig/interface/web/remote/index.php on line 14
    [Thu Jul 13 07:28:48.270736 2017] [fcgid:warn] [pid 6777] [client SOURCE_IPADDR:53430] mod_fcgid: stderr: #0 {main}
    [Thu Jul 13 07:28:48.270730 2017] [fcgid:warn] [pid 6777] [client SOURCE_IPADDR:53430] mod_fcgid: stderr: Stack trace:
    [Thu Jul 13 07:28:48.270678 2017] [fcgid:warn] [pid 6777] [client SOURCE_IPADDR:53430] mod_fcgid: stderr: PHP Fatal error: Uncaught Error: Class 'SoapServer' not found in /usr/local/ispconfig/interface/web/remote/index.php:14
     
    reason8, Jul 13, 2017
    #45
  6. till

    till Super Moderator Staff Member ISPConfig Developer

    Please install the php soap module on the target server and then run the migration script again on the source.
     
    till, Jul 14, 2017
    #46
  7. reason8

    reason8 Member

    That worked, for the most part. Soap is installed.
    The script did a dry run and created the bash script.
    I copied it to the new server and ran it.

    Received the following error.
    ERROR 1049 (42000): Unknown database 'DB_NAME'
    mysqldump: Got errno 32 on write
    find: ‘/var/www/clients’: No such file or directory

    Should ISPConfig have created this clients directory? Or am I supposed to?
    I don't remember doing this with opensuse but it's been a long time since I did a free install.
     
    reason8, Jul 15, 2017
    #47
  8. Taleman

    Taleman Well-Known Member HowtoForge Supporter

    If you did dry run of the migration script, you should not run the generated command file on the target server. Dry run does not create the users and databases on the target server, so the command file is not going to work.
     
    Taleman, Jul 15, 2017
    #48
  9. reason8

    reason8 Member

    Gotcha. That makes sense.
    However, now my target apache2 will not start (after the dry run only).
    Apache error is : AH00016: Configuration Failed
    I'm still learning ubuntu (as I already knew opensuse).
    I am not sure what happened here.
     
    reason8, Jul 16, 2017
    #49
  10. till

    till Super Moderator Staff Member ISPConfig Developer

    Please restart apache and then look into the apache error.log file for the error message. The log file is /var/log/apache2/error.log
     
    till, Jul 17, 2017
    #50
  11. reason8

    reason8 Member

    Apache will not start.
    Log shows the following: AH00016: Configuration Failed
    I'm not sure what this means. In opensuse I had an httpd.conf.local file I could check.
    I have not created this in Ubuntu yet.
     
    reason8, Jul 17, 2017
    #51
  12. till

    till Super Moderator Staff Member ISPConfig Developer

    Such a file is not used or needed on Ubuntu.

    Take a look into the folder /etc/apache2/sites-enebled/, are there any symlinks beside the ones with default or ispconfig in the name? If yes, delete the other ones and then try to restart apache.
     
    till, Jul 17, 2017
    #52
  13. till

    till Super Moderator Staff Member ISPConfig Developer

    and please run the commands:

    apache2ctl -t
    apache2ctl -S

    and post their output, if the above does not fix the issue.
     
    till, Jul 17, 2017
    #53
  14. reason8

    reason8 Member

    There was sylinked files like you said. They seem to have been from my source web server.
    Deleting all files prefixed with 100-, 500- and 900- resolved the issue.
    I left the files with 000-.

    I think I just ran the dry version. Is this occurrence possible on dry run?
    Thanks.
     
    reason8, Jul 17, 2017
    #54
  15. till

    till Super Moderator Staff Member ISPConfig Developer

    Not as far as I know, but @Croydon might know it better. I'll ask him to post his opinion on this. Just to be sure before you proceed, please login into ispconfig on the target. Is it still 'empty' or do you see sites and mail domains there? If there are sites and clients, then you probably run in real mode already and we have to ensure with the next run to use the right mode so you don't have all sites twice.
     
    till, Jul 17, 2017
    #55
  16. reason8

    reason8 Member

    No, it's not empty. I'm wondering if I ran the actual migration instead of the dry run the last time.
    But for some reason, if I did, the process did not complete. Which is why I could not start apache on target server.
     
    reason8, Jul 17, 2017
    #56
  17. till

    till Super Moderator Staff Member ISPConfig Developer

    Just to be sure, take a look at the jobqueue in the monitor module in ispconfig on the target, when this is empty too, then everything is fine. But I don't have an explanation for the vhost files then, but maybe Croydon has :)
     
    till, Jul 17, 2017
    #57
  18. reason8

    reason8 Member

    There is nothing in the Job Queue.
     
    reason8, Jul 17, 2017
    #58
  19. reason8

    reason8 Member

    Hello and good morning.
    I am holding off on the migration until I hear something back from you.
    Thanks.
     
    reason8, Jul 18, 2017
    #59
  20. till

    till Super Moderator Staff Member ISPConfig Developer

    Oh, sorry. When the job queue is empty and there is no data in ISPConfig, then the first runs must have been dry runs indeed. So you can proceed with the real migration.
     
    till, Jul 18, 2017
    #60
Page 3 of 5

Share This Page