Kolab 3.4: Disaster

[Matthias Busch]

Hi,

I dont know why, but I have a debian x64 7.8 install with kolab 3.3 
upgraded to 3.4, which does not show the apache or mysql errors you 
described.

single domain, handfull of users, active use of email and calendar via 
thunderbird, webmail and activesync.
there are a few smaller issues with which I'll have to post to the list 
looking for solutions, but nothing as horrific as yours.

size of caches (I am not a 100% sure that this is accurate as looking 
via mysql command prompt but since its the actual files, it should be?):

/var/lib/mysql/roundcube# ls -ahl
insgesamt 428K

-rw-rw---- 1 mysql mysql 8,6K Sep 30 22:08 cache.frm
-rw-rw---- 1 mysql mysql 8,5K Sep 30 22:08 cache_index.frm
-rw-rw---- 1 mysql mysql 8,6K Sep 30 22:08 cache_messages.frm
-rw-rw---- 1 mysql mysql 8,5K Sep 30 22:08 cache_shared.frm
-rw-rw---- 1 mysql mysql 8,5K Sep 30 22:08 cache_thread.frm
...
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_configuration.frm
-rw-rw---- 1 mysql mysql  13K Sep 30 22:08 kolab_cache_contact.frm
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_event.frm
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_file.frm
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_freebusy.frm
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_journal.frm
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_note.frm
-rw-rw---- 1 mysql mysql 8,7K Sep 30 22:08 kolab_cache_task.frm
-rw-rw---- 1 mysql mysql 8,6K Sep 30 22:08 kolab_folders.frm
...


apache is restarted every sunday (logrotate cron job with the fix of 
restart apache instead of "warm restart" which made it crash. it is a 
"known" issue, if I remember correctly with debian in general, not kolab)

I have yet to dive into the whole "what is linux doing with ram" stuff, 
but looking at:

# free -m
              total       used       free     shared    buffers cached
Mem:          8006       7861        144          0        162 6033
-/+ buffers/cache:       1665       6340
Swap:         1905          0       1905

and #ps aux | grep apache
root     29091  0.0  0.3 412328 27792 ?        Ss   10:25   0:00 
/usr/sbin/apache2.prefork -k start
www-data 29097  0.0  0.0 161056  4484 ?        S    10:25   0:00 
/usr/sbin/apache2.prefork -k start
www-data 29112  0.6  0.5 437152 41248 ?        S    10:25   0:14 
/usr/sbin/apache2.prefork -k start
www-data 29114  0.2  0.5 438504 42552 ?        S    10:25   0:05 
/usr/sbin/apache2.prefork -k start
www-data 29116  0.3  0.5 438348 42040 ?        S    10:25   0:06 
/usr/sbin/apache2.prefork -k start
www-data 29117  0.4  0.5 439872 43940 ?        S    10:25   0:09 
/usr/sbin/apache2.prefork -k start
www-data 29122  0.4  0.5 439928 43820 ?        S    10:26   0:09 
/usr/sbin/apache2.prefork -k start
www-data 29127  0.7  0.6 447668 51892 ?        S    10:26   0:16 
/usr/sbin/apache2.prefork -k start
www-data 29148  0.5  0.5 442556 46436 ?        S    10:26   0:12 
/usr/sbin/apache2.prefork -k start
www-data 29228  0.6  0.5 444772 48868 ?        S    10:27   0:12 
/usr/sbin/apache2.prefork -k start
www-data 29231  0.3  0.4 434916 38848 ?        S    10:27   0:07 
/usr/sbin/apache2.prefork -k start
www-data 29695  0.5  0.5 436848 41056 ?        S    10:42   0:06 
/usr/sbin/apache2.prefork -k start
www-data 30706  0.4  0.3 429400 32144 ?        S    10:56   0:01 
/usr/sbin/apache2.prefork -k start
www-data 30819  0.3  0.4 434908 38052 ?        S    10:57   0:00 
/usr/sbin/apache2.prefork -k start
www-data 30820  0.4  0.3 430520 31400 ?        S    10:57   0:01 
/usr/sbin/apache2.prefork -k start

I dont think my apache is behaving badly with memory either...

my monitoring (monit) shows:
system: 21% memory usage
apache 7% memory usage

I realise this doesnt help you. But somehow, 3.4 on debian 7.8 isnt 
broken like yours is for everyone... Maybe that information can help a 
bit...