Mecca/Pass1 Tables Help


Runs in Progress

RunTapeJobAction Status
codetime
sgie831a   fnsfu   7 hours
09950FD3143 138a21-Apr 12:05
09955FD3147 139w21-Apr 12:07

This table provided up to the minute information about the status of all jobs currently in the processing queues. It is typically quite reliable, but it should only be used as a guide for further investigation. In other words, if a job's status is this table seems strange (i.e. too much time spent in a given state) then the appropriate action is to log into the farm machine and investigate with CPS commands.

Run
The Run number of the first run on the tape.
The hyperlink goes to the Mecca Range page and centers the run range on the run in question.

Tape
The tape label for the job.
The hyperlink goes to the Whyme Database entry for the tape in question.

Job
The CPS job number.
The hyperlink goes Job Info page for the job in question.

Action Status Code
This field reflects the current processing state of the job. It's updated over the network from the farm systems. Therefore the status as reported by the view or show commands on the farm node should take presidence if the two states disagree.
s This job has been submitted and is waiting in the queue for further action.
h This job is on hold. See Pass1 CPS Commands for more information on hold.
w This job is in the CPS on-deck area. If this job is being staged then it's waiting to stage. Otherwise it's waiting to analyze.
i This job is currently staging it's input tape (FD) to disk.
rThis job has successfully staged it's data from tape to disk and is waiting to analyze.
a This job is in analysis.
o This job is currently staging it's output from disk to tape (FR).
c This job is in CPS clean-up. It's going through a brief rundown.
d This job is done. This state should last only until the logfiles are processed by factoids (at most one hour).
x This job has been canceled. See Pass1 CPS Commands for more information on cancel.
Code Colors
Green - Time spent in that state so far is appropriate. This does not mean that everything is alright. You will have to check the host node directly to say that.
Yellow - The time spent in the state is longer than expected. You might want to check the states on the host to confirm that the Mecca Access data is correct.
Red - The time spent in the state is too long. Check topx or xtopx on the farm host to ensure that the queue is processing data, and contact a Pass1 Czar if the queue is hung.

Action Status Time
This field reflects the time (CST/CDT) that the job entered the state reported by the Action Status Code field.


Runs Completed in the Last 36 Hours

Run #Iter.Vers. Tape #EventsJob #Status Oddpack
file
Log
Summary
ProcessedWritten
09948a4 SGI
FD31411039288950891 A 136D09948a.odd09948a.log
09945b4 SGI
FD31428375681455 B 125ES09945b.odd09945b.log
09949a4 AIX
FD31421066543974025 C 156P09949a.odd09949a.log

This table provides a
Mecca Range like list of jobs completed in the last 36 hours.

Run #
The Run number of the first run on the tape.
The hyperlink goes to the Mecca Range page and centers the run range on the run in question.

Iter.
This letter is a count of the iterations or attempts at processing of the tape in question.
The hyperlink goes to the Iteration History page.

Vers.
This field list the PassOne code version number and platform used in processing this job.

Tape #
The tape label for the job.
The hyperlink goes to the Whyme Database entry for the tape in question.

Events (Processed and Written)
These columns list the number of events read from the input tape (Processed) and the number of events written to the output tape. We typically expect about a 3-10% difference.

Job #
The farm queue and CPS job number.
The hyperlink goes Job Info page for the job in question.

Status
This field relates the completion status of the job in question.
Status Code Key
D = Job completed without any errors.
P = Job completed with one or more non-fatal errors.
E = Job terminated with one or more fatal errors.
ES = Job terminated with fatal error(s) and the tape has been resubmitted
The hyperlinks on the P and E codes go to the Iteration History page.

Oddpack File
This hyperlink takes you to an set of postscript oddpack summary plots. This set of plots is similar to the RECON plots from data shifts.

Log Summary
This hyperlink takes you to a summary of the job's logfile.


Runs in Range 9973 to 9979 (Mecca Range)

Run #Iter.Vers.Tape #EventsJob #Status
Pass1
Skim1
More Information
ProcessedWritten
9973b4 SGIFD31651029569935084A 140DK9973a.odd9973a.log
9974a4 SGIFD31661109689100195D 139PK9974a.odd9974a.log
9976a + 1
FD31671045105E 135ESU21-Apr 18:40
9977a4 AIXFD316835233272C 181EU9968a.odd9968a.log
9978

FD31701057573B 177SU21-Apr 18:47
9979

FD31711058663
N Uphoton data set 14 no mu1B

This table is the main interface to the official processing record for all tape to be processed in PassOne. From here you can:

This table should be used to choose runs for submission to the farm queues.

Run #
The run number of the first run on the tape.
The hyperlink centers the Mecca Range table on that run.

Iter.
This letter is a count of the iterations or attempts to processing this tape. The +1 appears for jobs with code ES to indicate that the current submission will add one iteration after completion.
The hyperlink goes to the
Iteration History page.

Vers.
This field list the PassOne code version number and platform used in processing this job.

Tape #
The tape label for the job.
The hyperlink goes to the Whyme database entry for the tape.

Events
Events for codes N and S.
Number of events on the raw data tape (FD) from the Whyme database.

Processed and Written for codes D, P and E.
These columns list the number of events read from the input tape (Processed) and the number of events written to the output tape. We typically expect about a 3-10% difference.

Job #
The farm queue and CPS job number.
The hyperlink goes Job Info page for that job.

Status
Pass 1
This field relates the PassOne status of the tape in question.
Pass1 Status Code Key
N = Tape not yet submitted to any queue.
S = Tape submitted and currently in a queue.
D = Tape was processed without any errors.
P = Tape was processed with one or more non-fatal errors.
E = Processing terminated with one or more fatal errors.
ES = Processing terminated with fatal error(s) and the tape has been resubmitted
The hyperlinks on the P and E codes go to the Iteration History page for the tape.

Skim 1
This field relates the Skim1 status of the tape in question.

Skim1 Status Code Key
U = Not yet skimmed.
K = Skimmed.
The hyperlink goes to the Skim Range page, centered on that tape.

More Information
For code N
Level 1 Trigger - First level or Master Gate trigger from Whyme database. Hyperlink goes to the Whyme database entry for that run.

For codes ES and S.
Submission data

For codes D, P and E.
Oddpack File - This hyperlink takes you to an set of postscript oddpack summary plots for the latest iteration. This set of plots is similar to the RECON plots from data shifts. Below are descriptions of each column and field:
Log Summary - This hyperlink takes you to a summary of the latest iteration's logfile.


Run Iteration History

Run #Iter.Vers.Tape #EventsWhyMe
events
JobOddpack
file
Log File
09841
b
4 SGI
FD304710392271039227A 12909841b.odd09841b.log
Failure Modes
  • Completed without Error.
  • 09841
    a
    4 AIX
    FD30478798601039227T 16209841a.odd09841a.log
    Failure Modes
  • Too few events processed
  • No C!!! completion message
  • Output tape error on fnckm fnckm19
  • Failed to allocate output staging disk.
  • This table is intended to keep track of all attempts to process a given tape. In addition to the standard tape information it also list the number of events and Whyme events side-by-side and it give a a full accounting of all existing error conditions.

    Run #
    The Run number of the first run on the tape.
    The hyperlink goes to the
    Mecca Range page and centers the run range on the run in question.

    Iter.
    This letter relates the iteration of the row. Each iteration of the tape should be represented in this table.

    Vers.
    This field list the PassOne code version number and platform used in processing this job.

    Tape #
    The tape label for the job.
    The hyperlink goes to the Whyme Database entry for the tape in question.

    Events
    This columns list the number of events read from the input tape.

    Whyme Events
    This columns list the number of events recorded in the Whyme database for this run. This number should agree with the last column for a successfully completed run.
    Job #
    The farm queue and CPS job number.
    The hyperlink goes Job Info page for the job in question.

    Oddpack File
    This hyperlink takes you to an set of postscript oddpack summary plots. This set of plots is similar to the RECON plots from data shifts.

    Log Summary
    This hyperlink takes you to a summary of the job's logfile.

    Failure Modes
    This field lists all failures modes found for that iteration. Green modes are non-fatal, Yellow mode are possibly fatal, and Red modes are most likely fatal.
    The hyperlink on each mode goes to the Failure Help page, which gives a description of the error and a recommended course of action.


    Job Info Tables

    This table provides detailed and up to the minute information about the progress of a particular job. This information is gathered as the job runs and is updated whenever something changes. During the execution of a job, the amount of information grows substantially. Most of the fields are self-explanatory, so this page only gives an overview.

    A typical table for a job in analysis will look something like this:

    Current Job Status
    Production Systemibme831d
    Job Number243
    Staging RequestedYes
    Job ProgressAnalysis Phase
    Time in State 3:46:23
    Run/Tape Info
    Input Tape FD3848
    Internal LabelFD3848
    Input Tape Drivefnckm30
    Requested Run10985
    Analyzed or Staged Run10985
    Staging Info
    Input Disk Requested01:08:16 12-May-1998
    Input Staging Disk/stageE871_01 on fnckm
    Input Staging Completed04:01:19 12-May-1998
    # of files input28
    Input Staging SuccessfulYes  
    Action Times
    Submitted20:15:00 11-May-1998
    Went On-deck01:08:02 12-May-1998
    Requested Input Disk01:08:16 12-May-1998
    Requested Input Tape01:09:05 12-May-1998
    Input Tape Mounted01:11:19 12-May-1998
    Finished Input Staging04:01:19 12-May-1998
    Finished On-deck04:01:22 12-May-1998
    Began Analysis07:30:36 12-May-1998
    Requested Output Disk07:32:29 12-May-1998
    Time spent On-deck 6:22:34
    Time for Input Mount 0:02:14
    Time spent Staging Input 2:53:03

    Current Job Status
    Gives information about the current state of the job, job #, etc.

    Run/Tape Info
    Lists what tapes are expected and what tape drives are used.
    Also gives results of checks of Run/Tape correspondence.

    Staging Info
    Gives information about the staging processes.
    The same info is collected for input and output staging.

    Action times
    Lists the time the job reached each checkpoint.
    If a job appears to be frozen this can sometimes be helpful in sorting out why.