Hard to tell. While the validators are hitting the disks to the max we also started the effort to bring Albert@Home online again - it also had to be shut down because of the server failure. This entails copying yet another dataset to the new server, affecting drive performance for validation. Anyhow, I'd say it's more on the order of hours than days...
Validation backlogs for FGRP4 and BRP* are already finished, most of these over last night.
S6BucketFU1UB validation is limited actually by disk I/O, i.e. there not much we can do to further speed it up. Validators are running and working as fast as they can, the problem is that results keep coming in at almost the same speed at which the validators work.
BM
Edit: this should get better while people run a lager share of FGRP4 work, and thus the incoming rate of S6BucketFU1UB declines.
LOL. :-) Cheers, Mike.
)
LOL. :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Ah here it is now running on
)
Ah here it is now running on both Einsteinmachines too. Fine fine :) First action on one..increasing the cache ^^
DSKAG Austria Research Team: [LINK]http://www.research.dskag.at[/LINK]
Nice one Bernd
)
Nice one Bernd
Just got home and now all my
)
Just got home and now all my machines are in the race again.
Hats off to Bernd and the rest of his crew for all the hard work!
Phil
RE: time for the
)
Darn, forgot about this one! Oh well, there's still a lot to do...
Einstein@Home Project
Not to be a jerk, or to
)
Not to be a jerk, or to dismiss your accomplishments, but is there any rough timeline on when the processing / validation backlog will be tackled?
I know how much time this sort of thing takes and appreciate your continued communication through a difficult and frustrating process.
Professional career: consultant/organizational researcher
Alternative job: Dodge-Chrysler-Jeep car site guy
Hard to tell. While the
)
Hard to tell. While the validators are hitting the disks to the max we also started the effort to bring Albert@Home online again - it also had to be shut down because of the server failure. This entails copying yet another dataset to the new server, affecting drive performance for validation. Anyhow, I'd say it's more on the order of hours than days...
Just keep an eye on the SSP...
Oliver
Einstein@Home Project
See server status
)
See server status page.
Validation backlogs for FGRP4 and BRP* are already finished, most of these over last night.
S6BucketFU1UB validation is limited actually by disk I/O, i.e. there not much we can do to further speed it up. Validators are running and working as fast as they can, the problem is that results keep coming in at almost the same speed at which the validators work.
BM
Edit: this should get better while people run a lager share of FGRP4 work, and thus the incoming rate of S6BucketFU1UB declines.
BM
RE: validation is limited
)
Does larger cache or disks with ssd buffers help in this case?
OK. Just wondering because my
)
OK. Just wondering because my stats have stayed the same since last week.
Professional career: consultant/organizational researcher
Alternative job: Dodge-Chrysler-Jeep car site guy