Regarding workunit id 71317639,the latest wingman timed out last Friday and the WU has yet to be sent out again. I know on some of the small projects, it can take a few days to a week for a WU to get resent but that doesn't fit this project. Did I just hit the "wrong WU at the wrong time" or is there a problem that's keeping this WU from getting resent?
Copyright © 2024 Einstein@Home. All rights reserved.
Is this a normal wait?
)
It does. :-)
Look at the Server Status page for "Oldest Unsent Result". The cause for the value being around one week (currently 6 d 22 h 51 m) is the locality scheduling. See the thread Excessive Disk Usage for some explanation. It contains also a link to the trac wiki.
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
RE: RE: ...but that
)
Thanks, basically, the locality scheduling effectively reduces the pool of possible users for a resend. Instead of being able to resend a WU to any same-platform user, it only has the same-platform users-with-the-same-set-of-data-files to resend it to. Oversimplified I'm sure but I think that basically explains why resends take so long on a main project like E@H. Locality scheduling looks like a good thing but everything has its pluses and minuses.