Resolved Failed Upgrade to Ubuntu 14

[GIDATO]

Hello

I am running Onyx, and I went through the upgrade as outlined in https://docs.plesk.com/en-US/17.0/a...ver-administration/distupgrade-support.74627/

After the dist-upgrade I went to run the distupgrade.helper.<OS.old-OS.new>_post.sh file and it was not there.

So, I have tried to follow the 'problem resolution' in the document above, but alas having lots of issues.

1. Apache not starting correctly. Had lots of 'garbled' modules (aclu, cloudflare, rigid, road) which I disabled as suggested. It did then start but does not link in anyway to proper virtual hosted sites.
2. Checked the dkpg - l | less and most plesk packages still refer to old ubuntu12. so moved on to enabling the repositories
3. I can't find plesk-installer

So.... my first questions are:
a) is there anyway to recreate the distupgrade.helper.<OS.old-OS.new>_post.sh file?
b) where is plesk-installer
c) is there anyway to rebuild the plesk system at this point and have apache running again, plus also have plesk admin running again?

If necessary I can rebuild the sites from backups!

Any help getting this server up and running again would be appreciated. If there is any information I can provide, please tell me what I need to upload here.

Thanks in advance

 

[UFHH01]

Hi GIDATO,

... relax ... all is still fine and you don't have to think about a backup right now.

Pls. make sure, that ALL source - lists at "/etc/apt/" now contain the NEW "trusty" source, instead of the "precise" definitions.
Have a look at for example to => #9 , where I posted some examples sources - lists.

After you made sure, that all sources now point to "trusty" sources, continue with:

apt-get clean
aot-get update
aptitude upgrade

Afterwards, pls. go through the apache - modules list and either upgrade them, or un-install and re-install them, so that the modules are the correct compiled ones for your current apache2 - webserver. Most modules are named with for example as "libapache2-mod-NAME", so it is pretty easy to find them with "apt-cache":

apt-cache search libapache2 | grep mod

In addition, it will help, when you use the Plesk repair utility after you are completely finished with the upgrades.

=> Plesk Repair Utility ( Plesk Onyx ( 17.0 ) - online documentation - Administrator's Guide )​

 

[GIDATO]

Thank you for your help.

I have checked the /etc/apt/sources.list and /etc/apt/sources.list.d/* and note that although everything is pointed to trusty, all are commented out within the sources.list.d directory, and certainly do not include the whole list like you have given in your examples at #9. Should I update the plesk.list file in that directory before doing the apt-get clean, etc.

 

[UFHH01]

Hi GIDATO,

my sources - lists - recommendations at => #9 is "to be used as it is" - no need to modify something, nor to fiddle around.

 

[GIDATO]

Hi

I'm taking one step at a time now.....

During the "aptitude upgrade" I got the following error several times.

/usr/bin/sw-engine: relocation error: /usr/bin/sw-engine: symbol CRYPTO_memcmp, version OPENSSL_1.0.0 not defined in file libcrypto.so.1.0.0 with link time reference

Is this something I need to worry about at this point?



And, when I used the apt-cache command, I get lots of modules. I tried upgrading the first one, and got the following:

$ sudo apt-get install --only-upgrade libapache2-mod-apparmor

Reading package lists... Done

Building dependency tree

Reading state information... Done

Skipping libapache2-mod-apparmor, it is not installed and only upgrades are requested.

0 to upgrade, 0 to newly install, 0 to remove and 3 not to upgrade.​

so, it looks like this wasn't installed. And, I'm a bit confused. Should I install all of the modules listed in the "apt-cache search libapache2" list, or only those that were already installed? i.e. should I drop the "--only-upgrade" option?

Thanks
David

 

[UFHH01]

Hi GIDATO,

/usr/bin/sw-engine: relocation error: /usr/bin/sw-engine: symbol CRYPTO_memcmp, version OPENSSL_1.0.0 not defined in file libcrypto.so.1.0.0 with link time reference

This is related to an OLD, not upgraded openssl - package on your server.
Pls. consider to think and react to current issues/errors/problems AFTER your dist-upgrade is finished ( incl. the Plesk repair procedure! ). During the whole process, you might definetly see ( possible ) error/warning - messages, which you are unclear about.

Should I install all of the modules listed in the "apt-cache search libapache2" list

Well, that's up to you. You are always able to disable modules that you don't need and this quite easily over "Home > Tools & Settings > Apache Web Server". The useless storage space consumption on your hard disk is very, very low, once you install all available modules and due to the fact, that you only have to click, to switch a module on or off, is a nice option in my opinion.

... or only those that were already installed? i.e. should I drop the "--only-upgrade" option?

Why complicated, when you have faster/easier ways ? That's why I recommended to use "aptitude", instead of "apt-get", because "aptitude" will offer you several ( possible ) solutions, in case of dependencies - issues for example.

 

[GIDATO]

Hello UFHH01

I've done all that, and on running "plesk repair -n" I still get the following:

/usr/bin/sw-engine: relocation error: /usr/bin/sw-engine: symbol CRYPTO_memcmp, version OPENSSL_1.0.0 not defined in file libcrypto.so.1.0.0 with link time reference
exit status 127​

I've also run the following:

$ apt-cache policy openssl
openssl:
Installed: 1.0.1f-1ubuntu2.21
Candidate: 1.0.1f-1ubuntu2.21
Version table:
*** 1.0.1f-1ubuntu2.21 0
500 http://gb.archive.ubuntu.com/ubuntu/ trusty-updates/main amd64 Packages
500 http://security.ubuntu.com/ubuntu/ trusty-security/main amd64 Packages
100 /var/lib/dpkg/status
1.0.1f-1ubuntu2 0
500 http://gb.archive.ubuntu.com/ubuntu/ trusty/main amd64 Packages​

and...

$ apt-cache policy libssl1.0.0
libssl1.0.0:
Installed: 1.0.1f-1ubuntu2.21
Candidate: 1.0.1f-1ubuntu2.21
Version table:
*** 1.0.1f-1ubuntu2.21 0
500 http://gb.archive.ubuntu.com/ubuntu/ trusty-updates/main amd64 Packages
500 http://security.ubuntu.com/ubuntu/ trusty-security/main amd64 Packages
100 /var/lib/dpkg/status
1.0.1f-1ubuntu2 0
500 http://gb.archive.ubuntu.com/ubuntu/ trusty/main amd64 Packages​

and I ran

$ldd /usr/bin/sw-engine | grep ssl
libssl.so.1.0.0 => /lib/x86_64-linux-gnu/libssl.so.1.0.0 (0x00007ff3c9872000)​

So, I am now stuck once again... I've done this command if it helps...

$dpkg -l | grep ssl
ii libcrypt-ssleay-perl 0.58-1build1 amd64 OpenSSL support for LWP
ii libflac8:amd64 1.3.0-2ubuntu0.14.04.1 amd64 Free Lossless Audio Codec - runtime C library
ii libflac8:i386 1.3.0-2ubuntu0.14.04.1 i386 Free Lossless Audio Codec - runtime C library
ii libgnutls-openssl27:amd64 2.12.23-12ubuntu2.5 amd64 GNU TLS library - OpenSSL wrapper
ii libio-socket-ssl-perl 1.965-1ubuntu1 all Perl module implementing object oriented interface to SSL sockets
ii libnet-smtp-ssl-perl 1.01-3 all Perl module providing SSL support to Net::SMTP
ii libnet-ssleay-perl 1.58-1 amd64 Perl module for Secure Sockets Layer (SSL)
ii libssl0.9.8:amd64 0.9.8o-7ubuntu3.2.14.04.1 amd64 SSL shared libraries
ii libssl0.9.8:i386 0.9.8o-7ubuntu3.2.14.04.1 i386 SSL shared libraries
ii libssl1.0.0:amd64 1.0.1f-1ubuntu2.21 amd64 Secure Sockets Layer toolkit - shared libraries
ii libssl1.0.0:i386 1.0.1f-1ubuntu2.21 i386 Secure Sockets Layer toolkit - shared libraries
ii libwavpack1:i386 4.70.0-1 i386 audio codec (lossy and lossless) - library
ii openssl 1.0.1f-1ubuntu2.21 amd64 Secure Sockets Layer toolkit - cryptographic utility
ii ssl-cert 1.0.33 all simple debconf wrapper for OpenSSL​

and see that libssl0.9.8 is still loaded. Might that be an issue? Is something still trying to access those versions (they do contain 14.04 in the name)

If so, what can I do to get them all on the same version?

Thanks
David

 

[GIDATO]

OK... I've found this...

$ apt-cache policy sw-engine
sw-engine:
Installed: 2.20.1-ubuntu1204.201610281305
Candidate: 2.20.1-ubuntu1404.201610281325
Version table:
2.20.1-ubuntu1404.201610281325 0
500 http://ftp.hosteurope.de/mirror/autoinstall.plesk.com/ubuntu/PSA_17.0.17/ trusty/all amd64 Packages
*** 2.20.1-ubuntu1204.201610281305 0
100 /var/lib/dpkg/status​

So, it looks like sw-engine needs to be updated.

I tried:

$ aptitude upgrade sw-engine
Resolving dependencies...
The following packages have been kept back:
sw-engine
No packages will be installed, upgraded, or removed.
0 packages upgraded, 0 newly installed, 0 to remove and 3 not upgraded.
Need to get 0 B of archives. After unpacking 0 B will be used.​

I see that it says '3 not upgraded'.

How do I get this upgraded?

 

[GIDATO]

OK - got that sorted with "apt-get dist-upgrade" now running repair.

 

[GIDATO]

Good Morning.

Last night I eventually got much of it working. I can now log onto Plesk Admin, and most sites are working (although there are still some Apache issues to sort out, but I think I can manage that).

During the repair, there were two issues, that I'm not exactly sure about"

1. There was an error during REPAIR INSTALLATION which resulted in

   Errors occurred while performing the following actions: fix apsc.
   WARNING!
   Some problems are found during fix credentials for APS controller database(see log file: /var/log/plesk/install/plesk_17.0.17_repair.log)

   The
   relevant parts of the log (I believe) are as follows:

   ===> Cumulative APS controller database (apsc) upgrade and repair has been started.
   ===> Cumulative upgrade and repair of APS controller database has been completed.
   Connection to APSC DB is not possible: Cannot load library: /usr/lib/sw/libmysqlserver.so.2.0
   APSC database has been registered successfully
   Connection to APSC DB is not possible: Cannot load library: /usr/lib/sw/libmysqlserver.so.2.0
   ===> Cumulative APS controller upgrade and repair (final stage) has been started.
   ===> Cumulative upgrade and repair of APS controller (final stage) has been completed.
   ERROR 1045 (28000): Access denied for user 'pma_LQoax3wVl75Y'@'localhost' (using password: YES)
   ===> Cumulative Plesk database upgrade and repair (revertable stage) has been started.
   ===> Preparing Plesk database upgrade (revertable stage).
   ===> Cumulative upgrade and repair of Plesk database (revertable stage) has been completed.

   . As I'm not sure what the APS Controller does, I'm not sure whether it needs to be fixed, and if so, what I need to do.

• There was a CRASH in REPAIR DNS:

  PHP Fatal error:  Call to a member function getPackageController() on null in /opt/psa/admin/plib/Repair/Dns.php on line 56
  exit status 255

I'm ignoring this, as it seems to be working.

Otherwise all good, thanks for your help, UFHH01.

 

[GIDATO]

Sorted the APS controller now using other posts on this forum

Not too worried about the DNS fix as I think my DNS is OK, but might be worth someone having a look at it to see why it might have failed.

 

[UFHH01]

Hi GIDATO,

plesk repair -n

I noticed by reviewing your answers, that you used the command "plesk repair -n". The correct usage would have been:

=> plesk repair ASPECT [OPTION] => ( for example! ) plesk repair all -n ( for a DIAGNOSTIC MODE - nothing will be repaired! ).

The ( my ) recommended usage would be: => plesk repair all -v -y
​

=> Plesk Repair Utility ( Plesk Onyx ( 17.0 ) - online documentation - Administrator's Guide )​