Einstein@Home offline August 14/15
Submitted on 8 Aug 2023 12:06:31 UTC
Due to necessary maintenance work on the electricity supply we need to shut down the E@H severs on Monday, August 14th at around 6 AM UTC. We plan to be back online on the following day, Tuesday, around 5 PM UTC.
Language
Copyright © 2024 Einstein@Home. All rights reserved.
Comments
OK. Tks.
)
OK. Tks.
ac.igor.gutierrez.dc-[at]-gmail.com
THANKS for the update as
)
THANKS for the update as always.
Thanks for the heads
)
Thanks for the heads up.
Just to clarify, does this include everything such as download servers, upload servers, message boards, etc.
If so, and to help those who aren't following the general news as closely as the more regular activity in Tech News, perhaps it should be mentioned there as well. There are bound to be questions like:-
Hopefully, everything will go so smoothly that you won't even need the 2nd day :-).
Cheers,
Gary.
Hi! - We will likely need
)
Hi!
- We will likely need to shut down the servers 1-2h earlier (4 AM UTC). Modifying the original news item will cause another round of client notification for everyone, I don't think that's worth it.
- The outage affects the whole data center, including all E@H servers (even those you don't notice from the outside, like the DB servers). Thus, yes, all of E@H will be completely offline.
- I'll keep you posted via the common BOINC forum at https://boinc.berkeley.edu/dev/forum_thread.php?id=10279
- This maintenance involves a lot of people from different organizations (university, our institute, external companies) that don't usually work together, and as far as I remember in the last 15 years, such a thing absolutely never went according to plans. In fact we gracefully shut down E@H to avoid a disaster as in Dec '21, when power to the E@H severs was cut hard and unexpectedly. The aftermath of this caused a handful of people working for weeks just to get hard- and software going again. I highly doubt that we could limit the downtime to one day, or should even try.
- Yes, suspending network activity on the client side for the downtime is certainly a good idea to avoid unnecessary backouts.
BM
There is a reference to this
)
There is a reference to this outage on the BOINC message boards, https://boinc.berkeley.edu/forum_thread.php?id=10279&postid=112519#112519.
That would be a good place to post any out-of-band news, in case of any unexpected snags.
Bernd Machenschalk wrote:....
)
Of course and thanks very much for the extra detail.
The original news item didn't really convey the nature or magnitude of the work. I was just trying to give you all a light-hearted comment (hence the smiley) wishing you all the best for an amazingly quick and successful operation.
Cheers,
Gary.
Einstein@Home is back online!
)
Einstein@Home is back online!
BM
Thanks, just noticed that you
)
Thanks, just noticed that you are and hopefully fully running again, and already succeeded to upload and report a whole bunch of WUs...
“We have normality. I repeat,
)
“We have normality. I repeat, we have normality. Anything you still can't cope with is therefore your own problem.”
BM
Long past the shutdown, my
)
Long past the shutdown, my Boinc Manager keeps alerting me to this event.
Why?
Mine too.
)
Mine too.