 |
 |
FYI: 2 new test passes should be finished tomorrow midday: /work/hallb/hps/data/engrun2015/tpass2.5 /work/hallb/hps/data/engrun2015/tpass2.6 tpass2.5 has a change that will likely fix errors #2 and #3 tpass2.6 adds #4 w/ new jar: surveyed ecal geometry None have fixes to problems with full field extrapolation that Omar mentioned. And fixing #1 relies on scipcomp at jlab, which will not happen until Tuesday. -Nathan On Sep 5, 2015, at 8:17 PM, Sho Uemura <[log in to unmask]> wrote: > 3. The first one (DataQuality.lcsim) makes a small text file with average collection sizes - useful for quick debug The second one (DataQualityRecon.lcsim) makes the root histograms. > > 5. That's my fault. I'll change it back. > > On Sat, 5 Sep 2015, Omar Moreno wrote: > >> On Sat, Sep 5, 2015 at 9:53 AM, Nathan Baltzell <[log in to unmask]> wrote: >> >>> Hi Everyone, >>> >>> The latest test finished yesterday (this one has full-field track >>> extrapolation to ecal): >>> /work/hallb/hps/data/engrun2015/tpass2.3 >>> >> >> ?Just FYI, I checked the full field extrapolation yesterday using the 2.3 >> files, and there seems to be an issue. I'm in the process of debugging it >> ...? >> >> >>> >>> 1. >>> Looks like some of the batch farm nodes don't have the links setup for the >>> latest gcc, >>> which the dst-maker needs. We hit this problem before pass1 too. I >>> submitted a CCPR. >>> >>> 2. >>> I noticed this error in every job's log files, but not clear which >>> hps-java process generated >>> it, nor when during the process, nor if it matters. Node memory limit is >>> set to 4 GB (same >>> as pass1), I will raise that too and see if this error goes away: >>> >>> java.lang.OutOfMemoryError: Java heap space >>> at >>> sun.misc.FloatingDecimal.readJavaFormatString(FloatingDecimal.java:1248) >>> at java.lang.Double.parseDouble(Double.java:540) >>> at >>> com.mysql.jdbc.ResultSetImpl.getDoubleInternal(ResultSetImpl.java:2517) >>> at >>> com.mysql.jdbc.ResultSetImpl.getDoubleInternal(ResultSetImpl.java:2489) >>> at com.mysql.jdbc.ResultSetImpl.getDouble(ResultSetImpl.java:2450) >>> at com.mysql.jdbc.ResultSetImpl.getObject(ResultSetImpl.java:4985) >>> at >>> org.hps.conditions.api.BaseConditionsObjectCollection.select(BaseConditionsObjectCollection.java:533) >>> at >>> org.hps.conditions.api.AbstractConditionsObjectConverter.getData(AbstractConditionsObjectConverter.java:126) >>> at >>> org.lcsim.conditions.CachedConditionsImplementation.getCachedData(CachedConditionsImplementation.java:20) >>> at >>> org.hps.conditions.svt.AbstractSvtConditionsConverter.getSvtGainCollection(AbstractSvtConditionsConverter.java:128) >>> at >>> org.hps.conditions.svt.AbstractSvtConditionsConverter.getData(AbstractSvtConditionsConverter.java:80) >>> at >>> org.hps.conditions.svt.SvtConditionsConverter.getData(SvtConditionsConverter.java:54) >>> at >>> org.hps.conditions.svt.SvtConditionsConverter.getData(SvtConditionsConverter.java:16) >>> at >>> org.lcsim.conditions.CachedConditionsImplementation.getCachedData(CachedConditionsImplementation.java:20) >>> at >>> org.hps.conditions.svt.SvtDetectorSetup.conditionsChanged(SvtDetectorSetup.java:82) >>> at >>> org.lcsim.conditions.ConditionsManagerImplementation.fireConditionsChanged(ConditionsManagerImplementation.java:122) >>> at >>> org.lcsim.conditions.ConditionsManagerImplementation.setConditionsReader(ConditionsManagerImplementation.java:69) >>> at >>> org.lcsim.conditions.ConditionsManagerImplementation.setDetector(ConditionsManagerImplementation.java:53) >>> >>> 3. >>> We have two steering files being run for data quality (same as pass1): >>> /org/hps/steering/production/DataQuality.lcsim >>> /org/hps/steering/production/DataQualityRecon.lcsim >>> Based on the outputs, looks like only one (the first one, I think) is >>> producing >>> something useful that is being saved to disk (root files). Is this >>> expected? >>> Probably a question for Matt. >>> >>> 4. >>> I made new detectors for the latest ECal survey with the y-shifts for top >>> and >>> bottom from Raphael. I tested them with hps-java, and hits are shifted as >>> expected, but did not make the lcdd files due to some error about gdml. >>> Can >>> someone generate them? HPS-EngRun2015-Nominal-v3* >>> >>> 5. >>> While testing, I noticed the latest trunk is now going to print >>> EventMarkerDriver >>> on every event. This makes an extra 100 GB in log files for the full >>> pass. But ok ? :) >>> >>> -Nathan >>> >>> ######################################################################## >>> Use REPLY-ALL to reply to list >>> >>> To unsubscribe from the HPS-SOFTWARE list, click the following link: >>> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1 >>> >> >> ######################################################################## >> Use REPLY-ALL to reply to list >> >> To unsubscribe from the HPS-SOFTWARE list, click the following link: >> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1 >> ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the HPS-SOFTWARE list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1