--- Begin Message ---
|
Subject: |
"PANIC: fatal region error detected; run recovery" |
|
Date: |
Wed, 20 Apr 2011 13:11:04 -0700 |
|
User-agent: |
Mutt/1.5.20 (2009-06-14) |
Hi all,
I'm mystified as to how I even got into this:
"PANIC: fatal region error detected; run recovery"
I see about running db_recover:
address@hidden:~/DB$ db4.8_recover -v
Finding last valid log LSN: file: 22 offset 120
Recovery starting from [21][657]
Recovery complete at Wed Apr 20 12:48:33 2011
Maximum transaction ID 80000007 Recovery checkpoint [23][28]
Which I guess looks good. But I'm still getting the error. I'm not
finding much help on Google (I swear this search engine is becoming
less useful by the day).
What now?
If it is relevant, I have been exploring sks dump; it seems
painless, taking 15-20 minutes or so to run, so if I can get this
ironed out, I think I should be able to offer nightly dumps.
Thanks!
--
David Benfell <address@hidden>
http://www.parts-unknown.org/
signature.asc
Description: Digital signature
--- End Message ---
--- Begin Message ---
|
Subject: |
corrupted database |
|
Date: |
Wed, 20 Apr 2011 15:38:48 -0700 |
|
User-agent: |
Mutt/1.5.20 (2009-06-14) |
Hi all,
Reviewing the logs, it appears my trouble began pretty close (I
don't know precisely) to when I did an sks dump. I'm now assuming
that's the culprit.
So several questions arise:
1) Can I use the dump I produced--apparently successfully--to
repopulate the database?
2) Should I? Or is there another way to recover from this?
(db_recover appeared to succeed, but apparently didn't.)
3) What is the proper way to do a dump so this doesn't happen?
I've shut down my keyserver until this is resolved; as near as I can
tell, it isn't doing anyone any good in its present condition.
Thanks!
--
David Benfell <address@hidden>
http://www.parts-unknown.org/
signature.asc
Description: Digital signature
--- End Message ---