BRP4 work availability
10 Feb 2023 8:10:43 UTC
Topic 229052
(moderation:
Hi everyone,
Please note that during the next 5 days there could be a small window of time where we don't have BRP4 work available. We already prepared a new batch of data that we'll move into our pipeline but we want to drain the current batch almost completely, due to post-processing requirements. Our current estimates put this window on Monday, Feb. 13th and we hope we can keep that window within a few hours max.
Cheers,
Oliver
Einstein@Home Project
Is there any estimate of how
)
Is there an estimate of how much BRP4 work there is left?
Thank you.
Try the server status page
)
Try the server status page ...
Thanks. I meant how much work
)
Thanks. I meant how much work in total there is left, not how many tasks are running/ready to send.
There used to be a progress counter but it hadn't been updated in years and it has disappeared from the Server Status page.
oops, sorry ... I
)
oops, sorry ...
I wonder if that can be estimated beforehand - might vary in depence of the results of successfully finished tasks.
Well, nevermind - I have no idea!
cheers
San-Fernando-Valley wrote:I
)
That's the same for the estimates of O3MD, FGRP5 and FGRPB1G which are displayed on SSP.
.
Do we get now new type of
)
Do we get now new type of work for BRP4/BRP4G? Current WUs are called something like guppi_57532_GBT820-01252_0008_0001_dms_3752_2040, before it used to be something like p2030.20200510.G55.05-03.86.N.b3s0g0.00000_2856. They also seem to need few CPU cycles more than the old ones (about 5% or so).
.
I believe guppi stands for
)
I believe guppi stands for Greenbank Ultimate Pulsar Processing Instrument.
A number of them have GBT in the work unit name also suggesting they are from the Greenbank telescope. We saw these on Seti@home at one point, although they were looking for signs of ET whereas Einstein are searching for Pulsars.
I have also started seeing some guppi work units with CAND in the work unit name so they may be from another radio telescope. Canada maybe?
MarksRpiCluster