1) Message boards : Number crunching : BHspin v0.05 (Message 849)
Posted 3 Dec 2015 by grn
Post:
Not sure which file you mean. Can you give me details?
2) Message boards : Number crunching : BHspin v0.05 (Message 846)
Posted 3 Dec 2015 by grn
Post:
I have 3 completed WUs that fail to report. I get the error: "Can't parse file info in scheduler reply; unexpected XML tag or syntax", followed by "No close tag in scheduler reply". Anybody seeing this?
3) Message boards : News : New ULX application (Message 834)
Posted 1 Dec 2015 by grn
Post:
I have a bunch of ULX WUs running at the moment. Their initial estimated time was 1 min, 2 secs. There are 16 running just now; all have been running for over an hour. The most complete has been running for over 2 hours and only 79% complete. There seems to be something seriously wrong with the initial estimate. The first batch of these WUs had an initial estimate of around 2 hours although; all ran for over the 2 hour estimate and many for over 2.5 hours and more. Just curious as I still have a reasonable number of ULX WUs with the very low initial estimate time.

I don't know how the scheduler deals with the work queue but I assume it must take in to account the expected time WUs in the queue have to run and presumably uses this to indicate whether the queue is full or further WUs can be accepted. I have a substantial number of WUs from other projects and no idea how long these ULX WUs will take to run. I will monitor the situation and will abort these WUs if I feel they will impact throughput for WUs from other projects.
4) Message boards : News : Longer work units (Message 767)
Posted 19 Nov 2015 by grn
Post:
Not sure if the latest lot of WUs I have are the longer ones but they certainly are long. Currently I have 29 WUs all with a deadline of 1st December 2015 and an estimated runtime of 563 hours. Unless this estimate is wildly out, I may as well abort them now as I won't make the deadline for any of them. With the WUs I have from from other projects, it is going to be about a week before I start on these so unlikely to make that deadline.

Is the estimate correct does anybody know. I have an i7 5960X processor.
5) Message boards : News : Longer work units (Message 691)
Posted 2 Nov 2015 by grn
Post:
Not sure if I've got the new WUs or not ...

You currently have 2 types in progress, the ones with a one week deadline are the "medium new" ones, the ones with a two weeks deadline the "brandnew" ones.

The one-week WUs have a somewhat longer runtime and a really bad estimated runtime (by mistake), the two-week ones run shorter but have a much higher estimated runtime. As your host probably has adjusted the correction factor to the one-week results, the two-week results will probably be listed with an estimated runtime of several days but will run less than one day.

The one-week results have a better time to credits ratio than the two-week results.

As soon as you have only brandnew (two-weeks) results left, the estimated runtime will slowly adjust to more realistic values I guess (but I'm not so experienced with the new estimate mimic of the 7.x core client so I might be wrong there).


Thanks for the explanation. I just signed up to this project and had no idea what to expect but just wanted to pass on my experience to date. Wasn't too bothered about the run times other than the effect it may have on other projects I'm subscribed too particularly as I had a large number of Universe@Home WUs with quite a short deadline but also had a large number of WUs from other projects that were also long running and all with a short deadline.

SInce I just changed my motherboard and CPU, going from a 4 processor/8 thread to an 8 processor/16 thread CPU helped cut the backlog down to size. I thought the low estimated run time would affect the WUs I get from the other projects I subscribe to.
6) Message boards : News : Longer work units (Message 688)
Posted 2 Nov 2015 by grn
Post:
Not sure if I've got the new WUs or not but they are coming in with an estimate of 1:25:29 on my i7-5960 but are running for around 14-15 hours on average. About 10 times the estimate or BOINC is doing a bad job benchmarking my system.







Copyright © 2024 Copernicus Astronomical Centre of the Polish Academy of Sciences
Project server and website managed by Krzysztof 'krzyszp' Piszczek