Yes, there were some server issues yesterday:
http://support.unamesa.org/message/view/TiddlyWiki.org/14956510
Everything should be back to normal now.
-- F.
Indeed. I've contacted the sysadmin, hopefully this will be fixed soon.
-- F.
"""
Turns out the machine's hard drive was corrupted. We're doing a restore
and hope to be back online soon (e.g. by tomorrow). Backups from
yesterday look good, so hopefully there will be no data loss.
"""
http://support.unamesa.org/message/view/TiddlyWiki.org/14956510#15063204
Please let us know if you experience any further issues.
-- F.
I understand UnaMesa (who host tiddlywiki.org on behalf of the
community) are working on recovering from backups.
Given a large number of us regularly take updates, I think at most we
will have collectively on "lost" only a few hours work. Still
annoying, but hopefully will be fixed soon!
--
Paul (psd)
http://blog.whatfettle.com
Yes - it seems the wrong backup had been restored. The proper recovery
is in progress:
http://support.unamesa.org/message/view/TiddlyWiki.org/14956510#15093712
> Given a large number of us regularly take updates, I think at most we
> will have collectively on "lost" only a few hours work.
I'm hopeful that we'll get a full backup, without losing any work.
I have a local Git clone of the SVN repo (last updated Oct 3), so worst
case I push that up to restore the data.
-- F.
We have been working hard to restore everything back to normal. I'll
elaborate below. UnaMesa will follow up with more details, including the
steps they are taking to prevent similar incidents in the future.
In short:
* No code was lost.
* A few minor commit messages were lost.
* Some Trac data (ticket changes, wiki edits) might remain lost.
The most recent backups had been corrupted. The latest valid backups
were taken on Sept. 6. Luckily, I had a local Git clone of the
Subversion repository containing all files' latest state as of Oct. 3
and including the commit history starting from Sept. 11.
Thus about 50 commits (between Sept. 7-10) had to be combined into one a
single commit:
http://trac.tiddlywiki.org/changeset/10612
The files listed in that commit suggest that there were no important
commit messages during that period.
The remaining commits (Sept. 11 to Oct. 3) were then reapplied.
As for Trac, unfortunately we'll have to manually restore the tickets
and wiki edits from that four-week period.
For that we're resorting to using Google's cache and our feed readers'
archives of the Trac timeline feed. This record is not complete (e.g.
new and closed tickets are covered, but not all ticket changes), so any
suggestions on alternative avenues would be most welcome.
-- F.
Thanks very much for your efforts! Very much appreciated.
J.
On Oct 8, 2009 1:54 PM, "FND" <FN...@gmx.net> wrote:> The proper recovery is in progress: > http://support.unamesa.org/message/view/TiddlyWiki.org/1495...
-- F. --~--~---------~--~----~------------~-------~--~----~ This message is part of the topic "s...
Indeed - great to know no code was lost here.
I assume both systems are still in read-only mode for the recovery
process. I've asked the sysadmins to restore write access to SVN, as
that should be fully operational now. I believe we're aiming for Trac to
reopen by Wednesday.
-- F.
That's done now.
-- F.
Smashing, thanks.
J.