bu_cms_history/USCRunAnalysis

SiteMap (Historical BU CMS wiki main page)

This page describes interesting stuff I've seen in runs taken in spring 2008 at USC. Search by run number.

2008-10-09 (hazen, BU)

Complaints from the pit about backpressure (OFW) from DCC and HTR EE at ~ 50kHz with fragment sizes averaging 1300 bytes. Jeremey suggests runs: 

  Without tuning of ZS thresholds (2.4k events on average): 64749
  After ~tuning of ZS thresholds (<2k events) : 64789

Look at 64789 in DQM. Fragment sizes for HBHE are ~ 1300 bytes.

2008-04-15 (hazen, CERN)

Runs 40558 and 40558 are copied. Took ~200k events of local runs in pit with ~ 12 FEDS. No data corruption seen.

2008-04-14 (hazen, CERN)

Recent runs: 

  257093808 Apr 11 15:12 USC_040601.root   OK
  257947811 Apr 11 15:43 USC_040609.root   OK
  26030006 Apr 11 16:07 USC_040618.root    OK
     72056 Apr 12 18:31 USC_040715.root
     72052 Apr 12 18:33 USC_040716.root
     14872 Apr 12 18:43 USC_040717.root
     14790 Apr 12 18:49 USC_040718.root
     14811 Apr 12 19:12 USC_040719.root
  19779574 Apr 14 10:24 USC_040778.root   OK
  25664281 Apr 14 10:31 USC_040779.root
  64703830 Apr 14 10:37 USC_040780.root  Feds 700, 702, 703 EvN mismatch
  26239503 Apr 14 10:55 USC_040781.root  OK
  24456235 Apr 14 11:05 USC_040782.root  OK
  25848470 Apr 14 11:10 USC_040783.root  OK
  25466349 Apr 14 11:13 USC_040784.root  OK
  25961030 Apr 14 11:26 USC_040785.root  OK
  27127747 Apr 14 11:34 USC_040786.root  OK
  24295041 Apr 14 11:48 USC_040787.root  OK
  35688601 Apr 14 11:52 USC_040788.root  OK
  24325505 Apr 14 12:08 USC_040789.root  OK

"OK" above means the DIM summary plot looked OK. Basically it means EvN match throughout run.

Run 40527. No problems seen. Dump raw data, EvN in sync at end of run.

Run 40539. Pedestal run, 1000 events. Fed 709 HTR 3 goes out of sync on Event 2. EvN 1 LRB err = 4b.

Run 40558. All data copied. EricDIM 1st 1000 events. Fed 704 HTR 5 has LRB=0xcb 

  - run 40558 - GT random triggers at 12KHz - 1.5-2KHz  daq rate- no daq prescaling - dies after 55K events: FED 704 goes   busy
  - cdaq stops, starts - HCAL reports "configured" - but we have a HCALsupervisor 0 = Faulty this happens twice
   HCAL sending busy to GT - GT sends resync and start - HCAL stays busy
  - destroy HCAL and try again

2008-04-13 (CERN)

Jason's bug hunt:

http://joule.bu.edu/~hazen/CMS/BugHunt.txt

Brief summary: 

  Runs 40551, 40558, 40564, 40565, 40575, 40665 have FED 709 busy
  Runs 40538, 40780, 40665 also have issues

2008-04-08

Run 40076. EricDIM on entire run: Plots . Overall rate average 300Hz.

Same BCN offset 7c from HTR to DCC (DCC has larger number). 

  FED 703 HTR 1 some problem around EvN 150k
  FED 715 HTR 7 problem around EvN 230k

EricScan output: 

 DUMP
 EvN 00155472 FED 003 HTR 01 reason "LRB errors=cb"
 DUMP
 EvN 00155473 FED 003 HTR 01 reason "LRB errors=0b"


 DUMP
 EvN 00233249 FED 015 HTR 07 reason "LRB errors=cb"
 DUMP
 EvN 00233250 FED 015 HTR 07 reason "HTR EvN=0x540005 FED EvN=0x038f22"

Running EricDump now.

2008-04-05

High-rate tests. See https://cmsdaq.cern.ch/elog/HCAL/4624

Run 40093 FEDs 704, 705, 709 data have problems. 704 went BSY at the end... maybe this is why the run stopped. See EricDIM_run040093.pdf . (full run). 

  Fed 704 HTR 5 (EvN 0x000013+)  
  Fed 705 HTR 3 (EvN 0x070478+)
  Fed 705 HTR 7 (EvN 0x03a514+)
  Fed 709 HTR 3 (EvN 0x061db3??)