In the past we were informed
14 Feb 2007 21:12:59 UTC
Topic 192428
(moderation:
In the past we were informed about what was happening with Einstein.
Now even although the project has been down for 5 or 6 days all I've got is
14/02/2007 20:13:19|Einstein@Home|Sending scheduler request: To fetch work
14/02/2007 20:13:19|Einstein@Home|Requesting 26 seconds of new work
14/02/2007 20:15:26|Einstein@Home|Scheduler request failed: HTTP internal server error
14/02/2007 20:15:26|Einstein@Home|Deferring communication for 27 min 55 sec
14/02/2007 20:15:26|Einstein@Home|Reason: scheduler request failed
14/02/2007 20:43:29|Einstein@Home|Sending scheduler request: To fetch work
14/02/2007 20:43:29|Einstein@Home|Requesting 26 seconds of new work
14/02/2007 20:43:32|Einstein@Home|Scheduler RPC succeeded
14/02/2007 20:43:32|Einstein@Home|Message from server: Server can't open database
14/02/2007 20:43:32|Einstein@Home|Deferring communication for 1 hr 0 min 0 sec
14/02/2007 20:43:32|Einstein@Home|Reason: project is down
14/02/2007 20:43:32|Einstein@Home|Deferring communication for 2 hr 34 min 37 sec
14/02/2007 20:43:32|Einstein@Home|Reason: project is down
What a shame!
Warning! This post contains atrocious spelling, and terrible grammar. Approach with extreme edginess.
It's not the speed, but the quality - Until I get a faster computer
In the past we were informed
)
What I've seen over the past week or so is that the server seems to be connection limited -- I'd get a 'server is down' when reporting a result, and then if I retried even a couple of minutes later, the upload would succeed. Same for downloads.
Clearly there is a constraint we've not encountered before.
Also, as you note, for whatever reason, status information is not being provided.
Similarly, when other problems occur (not just the 'narrow piple syndrome' I noted above), we are also left without information. At the moment, it seems there may be a database problem which is delaying validation. This seems to be something new for today (though I recall something like this about a week ago prior to a temporary (and unremarked upon) outage.
For me, the travails of the past two months have simply meant I reallocate CPU cycles to projects with higher current reliability. Sort of turnabout for me -- what drove me to add Einstein to my project list 21 months ago was reliability problems over on Seti. I've since added four other projects, and while Einstein has the largest share of total work done by me, it is now my 4th ranking project over the past two months.
There are other threads
)
There are other threads talking about this, and there have been some answer in some of them. Posting new threads on information we already has just confuses the issue more.
Part of the issue is a know BOINC database side issue. The BOINC developers are trying to get this corrected, then sent down the pipe to the other projects, so they can implement the fixes, and get things more on track.
Please just try and be patient. We are just volunteers, with nothing to lose. We help a project out of the goodness of our hearts. The teams that run the projects are usually a skeleton crew, who all have other jobs (full time professors, etc.). So their time is also limited in what they can do.
There are other worthy projects to crunch for. No one forces you to stay on one project, and it's actually encouraged to do several to help multiple causes, and keep your machines busy, if one does have an outage.
RE: There are other threads
)
You are absolutely right that we are just volunteers with goodness in our hearts (sic). That is why we could have a piece of information now and then! I am not leaving EAH despite the problems. After all I like seeing my pending credits growing by the day. (around 1600 by now)
Constantinos
Gravity increases significantly in Autumn, because apples fall in large numbers during that time!
I think a KEY here is to see
)
I think a KEY here is to see a post from one or more folks directly working on the project.
Not a post from a moderator who gleamed info from other places, or guesses by other posters.
My "pending" here is about 10 times what I ever saw before on this project.
Can't one of the actual project admins (not a forum moderator) give us a guess when the BOINC DB fixes might be installed? Are we talking a day, a week, or a month?
Moderators/guessors need not reply, and it's only a uninformed guess!
Matt L post Not a guess,
)
Matt L post
Not a guess, but an actual post by a BOINC developer explaining that there are database issues. If you read the whole thread, it does explain Einstein is part of this issue.
Sorry folks, we're still
)
Sorry folks, we're still having serious database problems. So far everything we are doing to solve them seems to have a temporary effect only, and the DB connections soon get unreliable again. We're working on this, but I don't have an estimation of when it will be solved.
BM
BM
Thank you for the post. This
)
Thank you for the post. This was by far one of the most reliable projects in BOINC until recently. I hope you manage to get things sorted out as B@A is chomping at the bit to do an AA here but the last 2 votes have narrowly gone to other projects due to stability issues so we have missed out on einstein for our AA's
P.S The more info we get the better.
RE: Sorry folks, we're
)
If possible, could you post this on the front page as well? The more visible location might actually manage to stop the whining for more than a day or two.
I was thinking that the
)
I was thinking that the additional load of boincsynergy having Einstein as POTM (project of the month) wasn't hurting Einstein. If the issue is a slow DB query, then reducing the load upon it might ease things up a bit. I'm setting my 7 hosts to NNW and will crunch elsewhere until the back end can handle the additional load.
I realize my efforts alone will only help about 0.0001%, but it's something I can do.
cu later
Good luck Admins
tony
RE: Sorry folks, we're
)
Bernd,
Would it be beneficial or detrimental if we would back off crunching for a little bit? If I suspend processing Einstein for a week or two will not hurt my computers, but would that help you and your team to try to get things on track? Or would it have a reverse effect, get things back in order, then all of a sudden all these machines come back online and cause more issues?
I think that is the information all of us would like to know.