Jay West <jwest at classiccmp.org> wrote:
In case no one caught my previous post on this, a few
days ago I moved =
the queue retention time back up to a reasonable value.
Thank you!
I suspect Michael Sokolov is testing this out for me,
his name servers =
(and thus email) have been down a few hours.
Do you really think that I, after having earned the trust of my circle of
friends (who all have accounts on my various servers, and use and depend on
various services hosted at my data centre) as a competent, reliable and
trustworthy professional system and network administrator for our Circle,
would deliberately shut it down, screwing all our users?! There must be
something seriously wrong with your thinking if you indeed thought so.
We (yes, we, not I, I'm not the only person at Harhan, and this server is
not the only one) had been down from about 02:30 UTC to 21:42 UTC (2005-03-08)
due to our enemies tampering with the SDSL line that connects our data
centre to the outside world. All of our servers were still running, but
the entire facility was isolated from the net for about 19 hours.
Apparently someone tried to wiretap the SDSL line and broke it in the process.
Time domain reflectometry indicated a problem in the line somewhere between
our facility and the local telco's CO from which the SDSL line is served.
The techs at the SDSL NOC assumed it was an open circuit, but I know this
was not the case, since during the downtime there was still *some* signal
coming from the SDSL line, it was just apparently too distorted for the
DSLAM and the CPE to sync up. So the problem detected by TDR, which they
dispatched a local telco tech to fix, was NOT an open, but something else.
I very strongly suspect foul play.
The fix took so long because the ISP apparently considers business SDSL
customers not worthy enough to fix in the middle of the night, so even
though I reported the problem last night, it wasn't until this morning EST
that they even looked at it, and then they had to have the local telco
dispatch a technician, and the tech had to get to it... I'm glad they fixed
it, but it's too bad I was at school at the time so I didn't catch the
telco tech and thus wasn't able to ask him what he found as the source of
the problem.
I'm now catching up with the backlog of mail...
MS