Hello!
I've encountered a really wierd problem with Midgard.
I was editing stuff with Nadmin (1.4.1) and used
a user account that is member of sitegroup's admin group.
For some non-obvious reason the following "things" occured:
1. All images stored to a page record vanished.
2. A number of subpages vanished (directly under root page).
3. Every snippet of the particular sitegroup vanished.
There was approx. 30 minutes between each of these events.
MySQL is running on another server (yes, two servers,
one serving Apache and Midgard, other MySQL).
I went to check the Midgard database (single db, no replication)
and found that the records had truly been erased. Then I
went to check the repligard table and all the repligard records
were looking good (everything in order, including "action" columns,
no "delete" actions).
Then I went to see if there's some scheduled replication/backup/etc.
tasks in crontab - nope, none of those were running during
the record disappearings. There's backup tasks running through
the night though.
Apache logs didn't give any hints. Only errors were due to not
found records (vanished images, pages and snippets).
Unfortunately MySQL was not logging during the events. :(
I'll turn the logging on and see if disappearings continue,
maybe I'm smarter after that.
In the mean time, has anyone encountered anything like this ?
I'll be greatfull for any advice where to look.
I'm no Linux-guru mind you. ;)
Best regards, and those of you who'll see Bergie and guys,
try to drink them under the table ('cause I can't)... ;)
//Henri Kaukola
--
Henri Kaukola [EMAIL PROTECTED]
Consultant Tel: +358-20-198 6037
Nemein Oy http://www.nemein.com
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
