[sf-lug] Jim Stockford (and/or others?): Do you have old list emails?

[Rick Moen]

Quoting Michael Paoli (Michael.Paoli at cal.berkeley.edu):

> Yep, true that for that outage, and guess I may well have been
> misremembering some much earlier shorter outage (days or so, or
> I thought, quite possibly incorrectly, there was a failure, a
> restore, and some (slight) bit of (e.g. day(s)) archive emails
> lost.  

Pretty sure not.

There was a fairly lengthy -- for West Menlo Park --
winter-storm-related PG&E outage in Dec. 2005 not long before I set up
sf-lug at linuxmafia.com in preparation for the initial (2005-12-26)
postings, but that wasn't long enough to exceed any SMTP delivery
timeouts.

Come to think of it, half a year later in summer 2006, there was a
hapless aDSL outage for something like most of a week that was basically
an AT&T screwup at the time we moved from 2033 Sharon Road (the
temporary rental house, a blocks from here) to 1105 Altschul Avenue (my
original family house where I've lived once more since 2006).  

I had carefully arranged for same-day cutover of my Raw Bandwidth aDSL
service on the day we moved, and supposedly AT&T (the landline provider)
was also ready to go.  But we moved... and there was no dialtone, and
Raw Bandwidth reported that AT&T was unable to activate our DSL at the
new address because the line was still in the name of the tenants who'd
just moved out.  So, despite all the planning, I had to put up with
quite a few days of unplanned outage anyway, and couldn't do a blessed
thing about it except wait.

Obviously, anyone who attempted to send linuxmafia.com mail during the
AT&T-caused downtime would have gotten 45x tempfail delivery status
notifications for a few days and then (if that went on too long) a 550
permanent failure one.  I'm pretty certain anyone who's mail was then
undeliverable either re-posted it after AT&T caught up with reality or 
-- after this many years -- no longer cares.

Some years _before_ 2005, maybe around 2002-3, I believe I had at one
point a sudden server hard drive death and, at that time, _did_ have to
revert to a several-day-old backup and thus lose a few days' traffic on
one or more linuxmafia-hosted mailing list.  And that takes us, gosh,
almost two decades back.

The current server hard drives date back to that period, which is
actually a little scary, but they've had continuous operation except for
successive motherboard failures, PG&E outages, the screwup when we moved
from Sharon Road, and the four months my server was kaput four four
months when I was laid off and then immediately there after had
catastrophic motherboard failure during a software upgrade.