A NICE JOKE!
6 Feb 2007 9:34:34 UTC
Topic 192393
(moderation:
Einstein@Home - Server Status
Einstein@Home server status as of 9:20 AM UTC on Tuesday, 6 February 2007 (updated every 120 minutes).
The Einstein@Home main server has been continuously up for 238 days 12 hours 25 minutes.
Gravity increases significantly in Autumn, because apples fall in large numbers during that time!
A NICE JOKE!
)
Why is that a joke?
The main server never failed. It was several servers around the main one that failed. The main database server stayed alive throughout all the outages.
Actually Pooh Bear is not
)
Actually Pooh Bear is not quite correct as it is the Project Server and not the Database Server that is referred to as the "Einstein@Home main server". Bruce has pointed this out previously in his reply to Alinator in this thread.
In a number of places Bruce has described the details of the various servers that make up the total project. If you haven't seen these messages before or don't remember what he posted, click on Bruce's name and then follow the link to all his message board posts and you will perhaps have a bit of an insight into the complexities of the server farm that runs the project and how much Bruce has posted not only about problems with the servers but also in the name of general user help as well. Please remember that a group of physicists and not a specialist server management team is applying the rubber bands and band-aids that is holding all this together :). I'm certainly prepared to concede that they are doing a pretty good job despite all the glitches over the last couple of months.
As an example of the detailed information that Bruce has provided, take a look at this particular message as to the cause of the "disk crashes" that required the rebuilding of the RAID Arrays.
We should be prepared to cut them a bit of slack when they have to deal with those sort of issues.
Cheers,
Gary.
Gary, Thanks for remining
)
Gary,
Thanks for remining us of that thread/post. It's interesting, because SETI have also recently been suffering from server/filesystem problems. In that case, the finger of suspicion is being pointed at an old, tired server, and they're trying to rustle up donations (cash or hardware) to replace it.
Eric Korpela has made it clear (here) that he wouldn't welcome WDC drives in his RAID, because of firmware concerns: yet Bruce Allen seems happy to use them. Is there any information-sharing between projects on issues like this - and if not, could we encourage it?
It's all in the semantics.
)
It's all in the semantics. Everyone is correct. But Constantinos is substantively correct... if you can't upload or download or get new work or report work then the system is down from the point of view of the contributors. And we feel the "pain".
Having said that let me be clear that I cut the project plenty of slack and I'm not in the least upset over recent problems. It's just that implying on the home page that the project has been up since inception when we all know our boinc message logs say "Project is down" is misleading to newcomers and a bit laughable to me at least.
As an example of the detailed
)
As an example of the detailed information that Bruce has provided, take a look at this particular message as to the cause of the "disk crashes" that required the rebuilding of the RAID Arrays.
Thanks for the links to Bruce's tech info. Very informative.