___/ On Fri 15 Sep 2006 17:29:38 BST, [ Owen Winkler ] wrote : \___
Roy Schestowitz wrote:
Owen, I hope to see you back on wp-hackers. Your insightful
posts are sorely missed.
Uh, ok. :)
Yes, the server is pretty dead. I guess the RAID card fried a bunch of
the drives that were connected to it. I don't blame Unixshell for my
inability to restore backups I don't have, and I should have known
better than to trust only their snapshot system. Getting the new VPS
running is another question, however.
I know I can offer no advice that is genuine (and yet
unheard of). However, I have learned from some miserable
experiences that one should store all all files in 2/3
separate (and preferably distant) sites. There are tools
that automate much of this and a one-time investment of
time/money is worthwhile as it stays with you forever. To
speak from personal experiece (hope you don't mind the
rambling):
I primarily use rsync (the bandwidth-efficient way) and scp.
It's possible to make complete copies of the hard-drive
every night, along with a stack of snapshots to avoid
'contamination'. It helps me sleep better at night, as well
as use wildcards. *smile*
The problem (and biggest enemy) is that with complacency
comes greed to be lenient and too trustworthy, which is why
I *force* myself, using a calendar, to do particular backup
tasks (and I still get sloppy on occasions). The routine is
mostly scripted (as much as possible), but it needs sanity
checks (I'm doing one at this very moment). Procrastination
is an enemy and that's only ever realises if when you say
"if only I had been more careful, I would have lost 1 day of
work instead of a week, or a month". In my case, a disaster
last stroke two weeks ago when I broke my Firefox profile
(third-party extension). It took hours to restore and it was
overall lossy. Another possibility that I take advantage of
(in case of 'contamination' on the server) is mysqldump as a
cron job on the Web server, or a stack thereof.
I'm not sure what all I have backups of, specifically the code for
RedAlt. Not a lot has changed since the backup I know I have (from
March), but I'm not sure what the status of the function reference is
for the moment.
I've recovered the blog posts that weren't backed up from Bloglines,
where I hadn't read anything since the last backup. All 175 posts. :)
Be sure to check your browser cache for the very latest
posts. Another backup procedure worth having is a nighly
wget of the front page. There have been cases in the past
where recent items somehow got b0rked, so the snapshot
proved to be useful in recovery.
I'm hoping to get some things back online today and over the weekend.
Once again, use the misfortunes of someone you know to take the time to
verify the workings of your backup plan.
It happens to all of us. It's only a question of: 1) how
often; 2) how severe; 3) recoverable/lossy/irreversible; 4)
how we handle this emotionally. I think of data loss as
though it's a mild analogy of recurring illness or the loss
of a loved one... but data and blog posts are the least
terrible thing that can happen, so we must be lucky.
Thanks to everyone for the kind thoughts.
Good to hear from you again.
Best wishes and happy (data) recovery,
Roy
|
