Message boards :
Number crunching :
Head scratcher_estimate clock on this
Message board moderation
Author | Message |
---|---|
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
07oc12ab.4870.33786.438086664198.12.162.vlar_1 I just rebooted twice after 2 rounds of Windows Updates and found this MB crunchin away with a 14+ Hour estimate clock! What tha heck? It's at 02:08 elapsed now and estimate remaining is trying to recover/update by leaps and bounds- about 6-10 seconds:1 declination but still showing 07:+ hours remaining and only 24% progress after 2 hrs Plenty weird - it thinks it's an CPU AP Don't know what to make of it |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
CPU AP tasks take my i5 2.9 about 5.5 hours to complete, so you're not far off that. remember that it takes BOINC somewhere in the rage of 15 completed tasks to figure out what the average runtime for your system is .... so yes, runtime looks high, but it will fall quickly. EDIT: I should read before I type LOL, a vlr file should probably take your computer a little over 2 hours to complete. |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
Heh - no worries Brent. I had just discovered my first 2 hr MB an hour ago. Never saw any even that big before. Abt 1:45 max is all I'd *seen til today - and it wasn't even a vlar to boot. It's the Akv8c xx AVXxx anonymous app and P Lasso showed pretty normal cpu usage These vary between 19-24% all the time I have Bitsum Highest and Game mode enabled with abovenormal priority and None affinity on all my SETI and Einstein tasks. They are pretty stable and consistent 95% of the time with those settings and not far off that without Lasso. Just a couple anomalies I haven't seen yet I guess - only been here abt 100 days - 8 Fails out of 15,460 tasks it sez. Anxious to see the stderr when it finishes (should say *if it finishes!) ;P = I'm running a Batch of 4 of these actually - all the same series - all but *this one have normal 2 hr clocks These things must be LOADED with Pulses or whatever - especially this Red-headed one - Sez it's ben running for 3 hrs now - at 82% and 30+ minutes togo. Hey Maybe it's a WOW! Heh heh. Standby -- then Stand Back! Should I call Raistmer? ;D |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
Here's the Finish: 3hrs 17 minutes runtime x 2hrs 42min CPU time =See anything weird? Task 07oc12ab.4870.33786.438086664198.12.162.vlar_1 Name 07oc12ab.4870.33786.438086664198.12.162.vlar_1 Workunit 1767949335 Created 22 Apr 2015, 11:36:34 UTC Sent 22 Apr 2015, 14:55:32 UTC Report deadline 14 Jun 2015, 19:55:14 UTC Received 23 Apr 2015, 22:26:43 UTC Server state Over Outcome Success Client state Done Exit status 0 (0x0) Computer ID 7459057 Run time 3 hours 17 min 18 sec CPU time 2 hours 42 min 49 sec Validate state Initial Credit 0.00 Device peak FLOPS 0.00 GFLOPS Application version SETI@home v7 Anonymous platform (CPU) Peak working set size 40.46 MB Peak swap size 52.09 MB Peak disk usage 0.05 MB Stderr output <core_client_version>7.4.42</core_client_version> <![CDATA[ <stderr_txt> Build features: SETI7 Non-graphics FFTW USE_AVX x64 CPUID: Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz Cache: L1=64K L2=256K CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3 SSSE3 SSE4.1 SSE4.2 AVX ar=0.008215 NumCfft=146007 NumGauss=0 NumPulse=50183457664 NumTriplet=67977339808 In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768 Windows optimized S@H v7 application Based on Intel, Core 2-optimized v8-nographics V5.13 by Alex Kan AVXxjf Win64 Build 2549 , Ported by : Raistmer, JDWhale SETI7 update by Raistmer Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.008215 Pulse: peak=6.361632, time=53.79, period=21.02, d_freq=1419078129.24, score=1.004, chirp=-1.8005, fft_len=2k Pulse: peak=3.38957, time=53.71, period=8.86, d_freq=1419078717.88, score=1.015, chirp=7.2028, fft_len=512 Pulse: peak=1.13354, time=53.69, period=1.633, d_freq=1419085494, score=1.044, chirp=13.338, fft_len=128 Autocorr: peak=18.0205, time=73.82, delay=5.5596, d_freq=1419083280.86, chirp=16.928, fft_len=128k Pulse: peak=2.598395, time=53.79, period=5.68, d_freq=1419086062.8, score=1.012, chirp=17.239, fft_len=2k Pulse: peak=9.722656, time=53.9, period=33.97, d_freq=1419078572.89, score=1.027, chirp=-29.928, fft_len=4k Pulse: peak=7.921866, time=53.74, period=22.37, d_freq=1419085662.41, score=1.02, chirp=-30.745, fft_len=1024 Pulse: peak=2.631691, time=53.74, period=5.592, d_freq=1419085664.79, score=1.002, chirp=-30.878, fft_len=1024 Pulse: peak=7.837197, time=53.9, period=22.23, d_freq=1419086755.49, score=1.061, chirp=-32.979, fft_len=4k Pulse: peak=8.014244, time=53.71, period=26.41, d_freq=1419079066.65, score=1.002, chirp=-39.214, fft_len=512 Pulse: peak=0.6875724, time=53.7, period=0.6717, d_freq=1419081757.78, score=1.002, chirp=41.082, fft_len=256 Pulse: peak=1.641711, time=53.74, period=3.141, d_freq=1419080215.64, score=1.016, chirp=41.281, fft_len=1024 Pulse: peak=3.130345, time=53.9, period=8.808, d_freq=1419083293.11, score=1.004, chirp=-77.844, fft_len=4k Build features: SETI7 Non-graphics FFTW USE_AVX x64 CPUID: Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz Cache: L1=64K L2=256K CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3 SSSE3 SSE4.1 SSE4.2 AVX ar=0.008215 NumCfft=146007 NumGauss=0 NumPulse=50183457664 NumTriplet=67977339808 In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768 Windows optimized S@H v7 application Based on Intel, Core 2-optimized v8-nographics V5.13 by Alex Kan AVXxjf Win64 Build 2549 , Ported by : Raistmer, JDWhale SETI7 update by Raistmer Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.008215 Pulse: peak=6.361632, time=53.79, period=21.02, d_freq=1419078129.24, score=1.004, chirp=-1.8005, fft_len=2k Build features: SETI7 Non-graphics FFTW USE_AVX x64 CPUID: Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz Cache: L1=64K L2=256K CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3 SSSE3 SSE4.1 SSE4.2 AVX ar=0.008215 NumCfft=146007 NumGauss=0 NumPulse=50183457664 NumTriplet=67977339808 In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768 Restarted at 3.84 percent. Pulse: peak=3.38957, time=53.71, period=8.86, d_freq=1419078717.88, score=1.015, chirp=7.2028, fft_len=512 Pulse: peak=1.13354, time=53.69, period=1.633, d_freq=1419085494, score=1.044, chirp=13.338, fft_len=128 Autocorr: peak=18.0205, time=73.82, delay=5.5596, d_freq=1419083280.86, chirp=16.928, fft_len=128k Pulse: peak=2.598395, time=53.79, period=5.68, d_freq=1419086062.8, score=1.012, chirp=17.239, fft_len=2k Pulse: peak=9.722656, time=53.9, period=33.97, d_freq=1419078572.89, score=1.027, chirp=-29.928, fft_len=4k Pulse: peak=7.921866, time=53.74, period=22.37, d_freq=1419085662.41, score=1.02, chirp=-30.745, fft_len=1024 Pulse: peak=2.631691, time=53.74, period=5.592, d_freq=1419085664.79, score=1.002, chirp=-30.878, fft_len=1024 Pulse: peak=7.837197, time=53.9, period=22.23, d_freq=1419086755.49, score=1.061, chirp=-32.979, fft_len=4k Pulse: peak=8.014244, time=53.71, period=26.41, d_freq=1419079066.65, score=1.002, chirp=-39.214, fft_len=512 Pulse: peak=0.6875724, time=53.7, period=0.6717, d_freq=1419081757.78, score=1.002, chirp=41.082, fft_len=256 Pulse: peak=1.641711, time=53.74, period=3.141, d_freq=1419080215.64, score=1.016, chirp=41.281, fft_len=1024 Pulse: peak=3.130345, time=53.9, period=8.808, d_freq=1419083293.11, score=1.004, chirp=-77.844, fft_len=4k Best spike: peak=23.62618, time=92.27, d_freq=1419084757.55, chirp=-23.484, fft_len=32k Best autocorr: peak=18.0205, time=73.82, delay=5.5596, d_freq=1419083280.86, chirp=16.928, fft_len=128k Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.122e+011, d_freq=0, score=-12, null_hyp=0, chirp=0, fft_len=0 Best pulse: peak=7.837197, time=53.9, period=22.23, d_freq=1419086755.49, score=1.061, chirp=-32.979, fft_len=4k Best triplet: peak=0, time=-2.122e+011, period=0, d_freq=0, chirp=0, fft_len=0 Flopcounter: 44675272589184.398438 Spike count: 0 Autocorr count: 1 Pulse count: 12 Triplet count: 0 Gaussian count: 0 Wallclock time elapsed since last restart: 6157.3 seconds 17:24:43 (4392): called boinc_finish </stderr_txt> ]]> |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
I don't see the 2.75 hours of work on a vlar as unreasonable. The 3.25 hours of runtime likely means you were working the computer doing other things. You were watching porn weren't you ... come on admit it :P |
Zalster Send message Joined: 27 May 99 Posts: 5517 Credit: 528,817,460 RAC: 242 |
come on admit it :P Hmm..That might explain the brown out we had earlier today... |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
@Zalster: Ha! If anybody had a Brownout, It woulda been from *you and your new GigaTitans over there; jamming 8 APs n MBs in all THREE of em at once and taking a Lunch break or sumthin! Youguyz probly have your own substation and backup gens anyway don't ya? I'm gonna hafta make good on my LEASH threat. ;) = No Brent, I did suspend everybody, exit Lasso and exit Boinc appropriately first before I ran those 2 Windows Updates and 2 shutdowns/restarts - which only took maybe 20 minutes tops including a BIOS 5-10 minute stop to confirm switches/settings. = As I said, I hadn't even *noticed any MB over 1:48 clock til today. So I *do think it's unusual. Could be just me tho First Encounter and all. This 3+ hr beast with a 14 hr opening clock on resume with 2hrs already elapsed was pretty big surprise. = Only real change I made in my BIOS was specifically selecting Digital DVI instead of Auto for my PEG - there's only one cable from GPU to my Mon. Display *is strikingly crisper; but c'mon that couldn't possibly have anything to do with this clock/runtime anomaly. |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
Run time 3 hours 17 min 18 sec CPU time 2 hours 42 min 49 sec This really shows the CPU was busy with something - did you check the running programs/processes? Do you use TThrottle with too low temperature set? Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
@BilBG - Yes I did look at processes etc as well as app and CPU behavior in Process Lasso. Did not see anything unusual; everything seemed *normal there (as I've come to define it). I'm still getting used to this Tool (Lasso) and what it's showing me, as well as interpreting those results and knowing what to do with them. I'm not a fan(no pun) of Throttling(down) a CPU or GPU to manage Heat - performance/production takes a hit there. I simply move fan speed up a notch. = My Board is an Intel branded DZ75ML-45K with i5 2500(non-K)3.3ghz Quad core 4 Thread(non-HT)with Intel stock Cooling solution(active setting) - Speedstep disabled but Turbo Boost enabled Cores run fairly cool (60c avg) under full load. I set all 4 cores active on both SETI and Einstein and do run Einstein GPU Binaries concurrently(but occaisionaly) along with my fulltime SETI work. The CPU usage on all apps at both sites rarely if ever exceeds 24% at any given time. 25% fulltime fan manages GTX 960 SC GPU temp at 65c running with 1.2060V and 1404mhz core and 1752mhz memory clocks at 85-99% load. As you can see, I set my core temps on both CPU and GPU below Factory Max listings; as I view those as extremes and believe in "ounce of prevention". = There may have been some tailing at play, post-Windows updates, but nominal at best; not enough to affect this app's runtime IMHO. The *test will be to find and observe a repeat performance and troubleshoot from there. |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
Here's the Finish: 3hrs 17 minutes runtime x 2hrs 42min CPU time Yes, the first section of the stderr shows the app getting to a pulse at chirp=-77.844 which would be over 90% progress on a .vlar. Then there's a restart which did not find the checkpoint file so started at the beginning again, and a third restart shown at 3.84 percent progress which then ran to completion. Basically the task ended up at about twice the normal run time because of that lost checkpoint, and the small progress just after the restart at beginning was what gave the extra long estimate to completion. Diagnosing how the checkpoint file got lost is probably impossible at this point, a Murphy's Law effect. I'd advise closing BOINC down before allowing Windows updates to run, though. Joe |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
Thanks for that Joe (and all the amazing work you do for us!);) Actually as is my habit, I *did carefully suspend everybody in BoincManager then shutdown/exit P Lasso AND Boinc (all stop), prior to *both Windows Updates and subsequent shutdown/restarts - yes, Both times. The only util I left running vs exiting was GPUz and confirmed my GPU had returned to Idle state. Boinc and P Lasso do start/load themselves on startup(my settings) and my GPU Nvidia Profile also loads but requires very minor adjustments (2 things) and they are non-critical. = The checkpoint thing, I've noticed in various/many stderr seems to lacking/missing 90% of the time. I'm not sure how to, when to or if, running a checkpoint diag flag from Advanced view is appropriate or could be helpful - or if I've got some setting amiss somewhere that's causing the error as such. I believe I set it to every 60 secs (somewhere) Are you aware of an appropriate repair or adjustment? Thanks |
Darth Beaver Send message Joined: 20 Aug 99 Posts: 6728 Credit: 21,443,075 RAC: 3 |
jbird i wouldn't worry i've got MB's times from 1 hr to 3 hrs 3 mins on the cpu's and 17 mins to 58 min's on the GPU's all normal , worry when they actually take 5 hrs then you have a problem |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
Thanks for the gracious note Glenn. The Q&A and feedback are part of the process and functions of the Forums here. A Help Desk dealing with a massive amount of variables. The runtimes ranges you just mentioned, has expanded my thresholds. I simply haven't/hadn't encountered such a thing(s) *yet. Since we're running similar apps on the same tasks with similar Hosts, your input is quite valuable. Thanks! I imagine your own *investigator is at work on that 10 hr AP I noticed in your Pending! ;) |
Darth Beaver Send message Joined: 20 Aug 99 Posts: 6728 Credit: 21,443,075 RAC: 3 |
Jbird when I do AP's they take any where from 10 hrs to 13 hrs so all normal for the cpu the Gpu one take 2 hr 30min to about 4 hrs max so all normal again Edit : I've noticed the AMD chip is not as fast as Intel on AP's done buy the CPU . My Intel Quad takes only 8hrs 30 mins to a max of about 11hrs 30mins . I don't have that one doing seti at the moment so you can't see that computers times at the moment (electric bill) It's a Intel Quad 8400 (I think , brain block at moment ) at 2.7 ghzs |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
I wonder if a cpl of tuned cmdlines would improve those Glenn. My i5 quad 3.3Ghz does the sse2 and avx versions in 5 hrs; and the NV versions in 1.5 hrs. I run them 1 at a time with a full core and full GPU - multiples of either don't interest me much. Your AMD is plenty strong and 16GB RAM is plenty too. |
Admiral Gloval Send message Joined: 31 Mar 13 Posts: 20277 Credit: 5,308,449 RAC: 0 |
I notice you are wondering about how long your tasks take. I looked through my current times. 4-5 hrs CPU MB. Up to 1.5 hrs GPU MB. Just for laughs. I have one of my Backup project CPU tasks. Name hadam3p_afr_yaqu_2013_1_009474066_2 Workunit 9555809 Created 29 Jan 2015 10:11:35 UTC Sent 29 Jan 2015 10:24:25 UTC Received 16 Feb 2015 3:57:20 UTC Server state Over Outcome Success Client state Done Exit status 0 (0x0) Computer ID 1340798 Report deadline 11 Jan 2016 15:44:25 UTC Run time 610,066.27 CPU time 543,007.80 Validate state Initial Claimed credit 12,546.97 Granted credit 12,546.97 application version UK Met Office HadAM3P-HadRM3P Africa v7.22 |
Darth Beaver Send message Joined: 20 Aug 99 Posts: 6728 Credit: 21,443,075 RAC: 3 |
My i5 quad 3.3Ghz does the sse2 and avx versions in 5 hrs; and the NV versions in 1.5 hrs. I run them 1 at a time with a full core and full GPU - multiples of either don't interest me much. The Intel is the older Core 2 Duo Quad and only has 2 gig of ram sort of a backup . Thanks yes it is plenty powerful I don't just use it for Seti thou . I do 2 Astropulse units per GPU and why the times are about 2 and harf hrs 2 is ok 3 is to much for my 650's |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
The odd ones(to me) in your AP batch is your 10hr runtime. stderr sez it ran under AVX (an Intel instruction set-incorporated into AMD processors in 2009 I believe). And this is on your 8 core AMD/16gigs RAM = Looks like you ran it as a single (vs 2x or more) Twice the runtime of my comparative sse2-AVX runs, just doesn't make sense(to me) = As far as MBs: my cuda50s(Lunatics x41zc app) avg 9min/20-25min at 0.04cpu + 0.5gpu config And the cpu versions(AKv8c_AVX app) are the 1:30,1:44(vlars) and 1:50 varieties along with the rare 2+hr like the 07oc tape that started this convo. - all running on their own cores at 20-25% cpu usage(the low one is due to the 2up cudas at 0.04cpu fraction-shared core) |
JBird Send message Joined: 3 Sep 02 Posts: 297 Credit: 325,260,309 RAC: 549 |
Gosh Admiral, I admire your patience and tenacity. Don't think I could stay onboard, much less "laugh" with 169 hrs per task! Fair winds and following seas, sir ;) |
rob smith Send message Joined: 7 Mar 03 Posts: 22204 Credit: 416,307,556 RAC: 380 |
CPDN is a project famed for its massive run-times, and as it all happens in the background it takes no time at all Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.