Topic: Oh dear - no server manager panels and now no email delivery

[avery]

OK so I have to admit my 6.01+update SME has been "a bit under the weather" for a while  now.

I lost access to the server manager first - it tries to log me in, then gives ....

Internal Server Error
The server encountered an internal error or misconfiguration and was unable to complete your request.

As the server was all configured OK, and everything else worked I didn't worry.....

Then it started not recovering from powercuts.  Its a 2HD raid, and hence I have to run fsck /dev/hda2 (& hdb2) several times at the file system repair console.  I can just about put up with this ...


However yesterday it stopped picking up mail (multidrop) - I'm getting the following emails ...

Cron <root@saturn> /etc/startmail
/bin/su: cannot set groups: Operation not permitted


Any thoughts on how I diagnose, and sort the problem - or is a upgrade and rebuilding a new 7.0b6 box more tempting??

[avery]

I'd overlooked this email before - this doesn't look good!!


Cron <root@saturn> run-parts /etc/cron.daily

/etc/cron.daily/logrotate:

error: error accessing /var/log/samba: No such file or directory
error: samba:1 glob failed for /var/log/samba/*.log
/etc/cron.daily/rkhunter:

/usr/bin/php: relocation error: /usr/bin/php: undefined symbol: XxqÙè¯ïƸ7ÅfÏ .1Y:«Aû©~h¯ñ>""¾"ÒW/áUY!äåf½Æïz"=v°xϦ.ùó=§fúMh£ß~§S/ã±S?صÁo£¿ybÃCwF,Z"ëjS.Qo^bcâ|e¡$öÚ.bYÖ¹fui¿"|'ï.¯(ÅsR¯ð}9ÌTæ3ØTßËuH_å?ÿü_offðÃ-,â°<Ë?ÿt¾à¦»~Ò0¥ãs"ö÷(îuyt
yµ¶ßÙ?÷¢8o8-gb½lðT<. ¿×æ7iH?=T>Û×V40ÿÊ_ 9ßi_Î&Í´ÇT²ð»@" Eü\'D·tîè9Â}^|±ZSÓÊáÏyøÄÕ³ "O?G1{ "l?r0i º»"ǺÅ^Åðé,XÒMñ±ÿý!ð.üIç°,'z¸y"ÿ'¾®sf!ÜØÅ÷àõëy FTî

[kruhm]

-carefully rebuild the raid
-/sbin/e-smith/signal-event post-upgrade
-/sbin/e-smith/signal-event reboot

NOTES:
-you shouldn't be going through powercuts. Buy a UPS.
-you shouldn't have to fsck anything after it boots into the OS as SME is based on the ext3: http://www.redhat.com/support/wpapers/redhat/ext3/

[avery]

Many thanks for the advice kruhm. The point about a UPS is well made!  

I'll follow the advice, but am puzzeled by the ext3 comment ... I'm just doing what the box is asking for at the recovery console!

I'm going to set up a second SME7, and backup info across two boxes, rather than RAID1, which seems a bit vunerable with a single point of failure.

Any other advice welcomed!

[avery]

Experimenting with a second new 7.0b8 box now.  
 
Anyone any thoughts on my current poorly 6.01 box issues???  

[kruhm]

- Boot the box
- Wait for the SME Server splash screen
- Hit CTRL-X
- Hit tab
- Enter the [image name] single
- enter with root password
- e2fsck -y /