 |
Procedure: |
SofI Pipeline Supervision
|
Keywords: |
NTT - SofI - Pipeline supervision |
Prepared/Revised: |
vivanov 2006-09-12 |
Reviewed: |
|
Released: |
|
|
Revision History:
2006-09-12 v1.0 V.D.Ivanov creation
2006-09-14 v1.1 V.D.Ivanov revision
2006-11-07 v1,2 V.D.Ivanov revision
|
Summary
This page contains information how to start/stop the SofI pipeline.
Normally, the pipeline is maintained by the data handling administrator
who monitors if the pipeline works normally.
However, the support astronomer and the telescope operator should be
able to execute basic operators such as starting and stopping the
pipeline in case of failure and emergencies, i.e. in the middle of the
night. This page is meant to help them in this task.
Procedure
0. General
The pipeline runs continuously. The data are transfered to and from
the pipeline machine automatically by permanently running
dataSubscribers (not to be confused with the dataSubscriber running
on the wg5off). The reduced data are located in
astro@wg5off:/data/reduced/yyyy-mm-dd
For more information see the
SofI pipeline page.
1. Starting the SofI pipeline and checking if it is working
The pipeline machine pipeline@wg5pl.ls.eso.org is located
behind the NTT console, its monitor, keyboard and mouse are on the
high shelve above the wg5off machine. The userid is pipeline, the
password is the usual.
- Open a terminal on the pipeline machine.
- Type: pipelineControl start
The output should be:
using DHS_DATA_BACKLOG /data/backlog
- Type: startDO sofi & (this is the Data Organizer)
The output is of the type:
...
18:39:20 [INFO] Frame /data/raw/2006-11-06/SOFI.2006-11-07T09:02:49.739.fits received
18:39:20 [INFO] Frame /data/raw/2006-11-06/SOFI.2006-11-07T09:02:49.739.fits processed and classified as IM_JITTEROBJ
18:39:20 [INFO] No RB created for /data/raw/2006-11-06/SOFI.2006-11-07T09:02:49.739.fits due to e.g. no reduction rule defined
...
18:39:20 [INFO] EndOfTemplate event for /data/raw/2006-11-06/SOFI.2006-11-07T09:08:52.218.fits
18:39:20 [INFO] ReductionBlock /data/redblock/SOFI.2006-11-07T08:57:59.303_tpl.rb created
...
- Type: startRBS pipeline.config & (the config file contains the startDRS command)
The output is of the type:
18:41:46 - RB `SOFI.2006-11-05T20:42:10.117_tpl' ready for execution
Reduction System is idle...
18:41:46 - Execution of RB `SOFI.2006-11-05T20:42:10.117_tpl' started
18:41:46 - RB `SOFI.2006-11-05T20:42:24.738_tpl' ready for execution
Reduction system is busy...
18:41:46 - ERROR: RB `SOFI.2006-11-05T20:42:10.117_tpl' failed:
(last o.k. app.: /midas/05SEP/monit/helper.exe)
stopped in Midas monitor
...
18:44:26 - RB `SOFI.2006-11-06T00:03:08.691_tpl' terminated o.k.
18:44:26 - RB `SOFI.2006-11-06T00:14:12.873' ready for execution
Reduction System is idle...
18:44:26 - Execution of RB `SOFI.2006-11-06T00:14:12.873' started
18:44:26 - RB `SOFI.2006-11-06T00:14:59.461' ready for execution
Reduction system is busy...
18:44:27 - RB `SOFI.2006-11-06T00:14:12.873' terminated o.k.
18:44:27 - RB `SOFI.2006-11-06T00:14:59.461' ready for execution
Reduction System is idle...
18:44:27 - Execution of RB `SOFI.2006-11-06T00:14:59.461' started
18:44:27 - RB `SOFI.2006-11-06T00:18:07.359_tpl' ready for execution
Reduction system is busy...
...
DO NOT BE AFRAID OF ERROR MESSAGES! The pipeline just does not reduce every type of data!
In addition, this command opens a xgterm that shows the output of the pipeline recipes.
- To check if the pipeline is working execute:
ps -eaf | grep dhs which should give you an output
containing three processes, - dhsSubscribe, vcsolac and
dhs, - similar to this:
pipeline 3560 1 0 Aug25 ? 00:02:35 dhsSubscribe -dhshost wg5dhs -dhsdata /data/raw -logpath /data/msg -polldir /data/msg -baddir /data/bad -backlogdir /data/raw -verbose 0 -where (INSTRUME!=MASCOT&INSTRUME!=VIMOS)|(INSTRUME=VIMOS&OCS.CON.QUAD>0) -id PIPE -backlogdir /data/backlog -run ~flowmgr/dfs/bin/linkToPipeline %s
pipeline 3596 1 0 Aug25 ? 00:00:05 vcsolac -dhshost wg5pl -dhsdata /data/lists/reduced_olas -baddir /data/bad -polldir /data/lists/reduced_olas -logpath /data/lists/reduced_olas -id PRO -verbose 0
pipeline 3631 1 0 Aug25 ? 00:02:47 dhs -use-threads 1 -dhsdata /data/reduced -logpath /data/msg -polldir /data/msg -baddir /data/bad -verbose 0
pipeline 20872 20738 0 16:17 pts/4 00:00:00 grep dhs
2. Stopping the SofI pipeline
- Type: stopRBS
- Type: stopDO
- Type: pipelineControl stop
3. Updating the calibration database
_oOo_