So... It appears to me that the list is up again.

Here's the short version of what happened.

First, the context.  This month, I'm doing my regular job for 4-8 hours a
day and spending a lot of the rest of my time working on the house we just
bought -- we plan to be in there (or at least out of here) at the end of
this month.  So, without any server troubles, I was already very busy.

About a week ago, maybe more, the main mccmedia.com server died.  It would
power up and do absolutely nothing... no POST, etc.  So I immediately bought
a new motherboard and disks, figuring that would be the fast way to get back
up.

The first problem was that Solaris refused to install on the new hardware.
There was various bitching about 64-bit stuff, but I couldnl't get it to
work.

I installed FreeBSD.  But even though FreeBSD uses the UFS file system,
there apparently is no way to read Solaris UFS disks on a FreeBSD system.

I bought more hardware, so that I could build a Solaris box and get the data
off the old disks.  For several days, it seemed that I could either have the
disks or the network, but never both at the same time.  This was...
difficult.

Yesterday, I finally got both machines operating and networking... then it
was a matter of recovering the data off the old disks, copying to the new
machine, playing with various configurations and so forth.  And if you're
reading this, it finally all worked.

The good thing about all this was that I knew that there were copies of all
of the data on two of the old disks -- a long time ago, I set up a cron job
that regularly copied all the Mailman files onto a backup disk.

The bad thing was that the active disk and the backup were both Sun UFS
disks.  Now I realize that I should have kept a backup on a Windows or other
common format machine... not on the only machine that can read that
particular file system.  So... guess what I'll be adding.

Please accept my apologies for the lengthy down time.  If I wasn't
remodeling at the same time, this would have been only a day or two.  But so
it goes...

I'll also take another look at how hard it might be to mirror the data on a
server elsewhere, so that we can fail over rapidly if something like this
happens again.

I'll sleep better tonight.

Nick

-- 
Nick Arnett
[EMAIL PROTECTED]
Messages: 408-904-7198
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l

Reply via email to