Checking backup (and other batch jobs)

I would like a program to run that checks on the status of batch jobs (eg backups). I would like a mail sent with different subject line for - job ran successfully - job generated errors - job appear not to have run I started to write something using a mixture of SEARCH -e- etc and MAIL but was starting to think someone must have written this better than I can do quickly Any suggestions A

In article <Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94>, "Andrew Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> writes: >I would like a program to run that checks on the status of batch jobs (eg >backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > >I started to write something using a mixture of SEARCH -e- etc and MAIL >but was starting to think someone must have written this better than I >can do quickly > >Any suggestions Our batch jobs do this by themselves. The only "external" check we have is a procedure that checks, wether a given job (name) is actually present. This procedure runs once a day and checks the existence of some important batch jobs. Regards, Christoph Gartmann -- Max-Planck-Institut fuer Phone : +49-761-5108-464 Fax: -452 Immunbiologie Postfach 1169 Internet: gartmann@immunbio dot mpg dot de D-79011 Freiburg, Germany http://www.immunbio.mpg.de/home/menue.html

"Christoph Gartmann" <gartmann@nonsense.immunbio.mpg.de> wrote in message news:equlev$98i$1@news.BelWue.DE... > In article <Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94>, "Andrew > Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> > writes: >>I would like a program to run that checks on the status of batch jobs (eg >>backups). I would like a mail sent with different subject line for >> - job ran successfully >> - job generated errors >> - job appear not to have run >> >>I started to write something using a mixture of SEARCH -e- etc and MAIL >>but was starting to think someone must have written this better than I >>can do quickly >> >>Any suggestions > > Our batch jobs do this by themselves. The only "external" check we have is > a > procedure that checks, wether a given job (name) is actually present. This > procedure runs once a day and checks the existence of some important batch > jobs. > > Regards, > Christoph Gartmann > > -- > Max-Planck-Institut fuer Phone : +49-761-5108-464 Fax: -452 > Immunbiologie > Postfach 1169 Internet: gartmann@immunbio dot mpg dot de > D-79011 Freiburg, Germany > http://www.immunbio.mpg.de/home/menue.html We do similar things as well. We have jobs that run that check that other jobs are submitted, daily jobs that search the error logs for any -E- messages etc. I also recently implemented a process that records the start and stop times of jobs, dumps that down to our intranet and then if the jobs are running outside of "normal" ranges or are late starting, alerts are issued. I only have a couple of critical jobs set for alerts though such as our pre-overnight backup which if it hangs up, can cause a bad day for a lot of people if not resolved quickly. I do miss the days when there was a team of operators actually paying attention to things. Andrew B

Andrew Butchart wrote: > "Christoph Gartmann" <gartmann@nonsense.immunbio.mpg.de> wrote in message > news:equlev$98i$1@news.BelWue.DE... > >>In article <Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94>, "Andrew >>Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> >>writes: >> >>>I would like a program to run that checks on the status of batch jobs (eg >>>backups). I would like a mail sent with different subject line for >>>- job ran successfully >>>- job generated errors >>>- job appear not to have run >>> >>>I started to write something using a mixture of SEARCH -e- etc and MAIL >>>but was starting to think someone must have written this better than I >>>can do quickly >>> >>>Any suggestions >> >>Our batch jobs do this by themselves. The only "external" check we have is >>a >>procedure that checks, wether a given job (name) is actually present. This >>procedure runs once a day and checks the existence of some important batch >>jobs. >> >>Regards, >> Christoph Gartmann >> >>-- >>Max-Planck-Institut fuer Phone : +49-761-5108-464 Fax: -452 >>Immunbiologie >>Postfach 1169 Internet: gartmann@immunbio dot mpg dot de >>D-79011 Freiburg, Germany >> http://www.immunbio.mpg.de/home/menue.html > > > > We do similar things as well. We have jobs that run that check that other > jobs are submitted, daily jobs that search the error logs for any -E- > messages etc. I also recently implemented a process that records the start > and stop times of jobs, dumps that down to our intranet and then if the jobs > are running outside of "normal" ranges or are late starting, alerts are > issued. I only have a couple of critical jobs set for alerts though such as > our pre-overnight backup which if it hangs up, can cause a bad day for a lot > of people if not resolved quickly. > > I do miss the days when there was a team of operators actually paying > attention to things. > > Andrew B > > I used to use the JAMS job scheduler which provided error handling, notification of failed jobs, etc. The notification was by e-mail which could be sent to a pager or a cellular telephone. It also handled dependencies so that if job B required that Job A complete successfully, job B was not started until Job A completed successfully.

On Feb 14, 4:30 am, "Andrew Black (delete obvious bit)" <andrew6e.deletet...@mail.black1.org.uk> wrote: > I would like a program to run that checks on the status of batch jobs (eg > backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > > I started to write something using a mixture of SEARCH -e- etc and MAIL > but was starting to think someone must have written this better than I > can do quickly > > Any suggestions > > A (Shameless plug!) Perhaps this would be useful for you: http://h71000.www7.hp.com/freeware/freeware60/notify/ It will watch a job for you and NOTIFY you when it's done and provide the completion status of the job. It also works if the job is deleted before execution. You can direct it to notify you via REPLY and/or MAIL. AEF

"Andrew Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> wrote in message news:Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94... >I would like a program to run that checks on the status of batch jobs (eg > backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > > I started to write something using a mixture of SEARCH -e- etc and MAIL > but was starting to think someone must have written this better than I > can do quickly > > Any suggestions > > A My batch jobs take care of themselves and record success/failure to a file that is scanned by another process which will email the logfiles of jobs failed or which reported errores in processing. But you could also use SET QUE/RETAIN and scan the batch queues to find what you seek (jobs failed, jobs that did not run, etc). There are many many ways to handle this depending on your need. Actually I do a bit of both as users can check their own jobs in the queue (with the RETAIN being set on the queues) and look at the status and the log files, from a screen I wrote for them. Syltrem

Another method (which isn't entirely foolproof) would be to define system-wide or cluster-wide logical names at various times within the backup job (e.g. define/system bckstat "Backup of DSA1: started at ''f $time()'" just before you issue the BACKUP command) and then pick up the status at the end of the job and redefine the logical then too. You can then either translate the logical names, write them out to a file and mail the file or even set up a job to translate the logical names on the fly and use a command procedure in a webserver script to generate a web page. Lack of logical names could be handled by a message like "Backup status not determined - check the queues" Steve Andrew Black (delete obvious bit) wrote: > I would like a program to run that checks on the status of batch jobs (eg > backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > > I started to write something using a mixture of SEARCH -e- etc and MAIL > but was starting to think someone must have written this better than I > can do quickly > > Any suggestions > > A

"Andrew Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> wrote in message news:Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94... >I would like a program to run that checks on the status of batch jobs (eg > backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > > I started to write something using a mixture of SEARCH -e- etc and MAIL > but was starting to think someone must have written this better than I > can do quickly > > Any suggestions > JAMS can do most of this just by installing it and configuring the e-mail addresses. If that's all you need it to do, the free license will work just fine. Go to http://www.mvpsi.com for details. The part that it might not be able to do is the "job appear not to have run". We can send you e-mail if a job doesn't start within x minutes of it's scheduled time, is that what you're looking for? If so, that would require some configuration, you would have to define the jobs and their "stalled time".

You will find my take on how to do this with DCL documented in this article: http://www.migrationspecialties.com/pdf/Using%20OpenVMS%20to%20Meet%20a%20Sarbanes-Oxley%20Mandate2.pdf Bruce Claremont www.MigrationSpecialties.com OpenVMS Stealth Marketing Squad

In article <Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94>, "Andrew Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> writes: > I would like a program to run that checks on the status of batch jobs (eg > backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > > I started to write something using a mixture of SEARCH -e- etc and MAIL > but was starting to think someone must have written this better than I > can do quickly I embed this in my batch jobs. An "on" statement captures jumps to a label which catches $STATUS and generates an email, then lets the batch job continue if appropriate. It would be trivial to add an email at the end of the job. Usually submit/notify serves this purpose instead. A batch job that didn't run? Missing from the email.

In article <Xns98D761BAFC4CEandrewblack1orguk@193.111.200.94>, "Andrew Black (delete obvious bit)" <andrew6e.deletethis@mail.black1.org.uk> writes: > I would like a program to run that checks on the status of batch jobs (eg > backups). I would like a mail sent with different subject line for > - job ran successfully > - job generated errors > - job appear not to have run > > I started to write something using a mixture of SEARCH -e- etc and MAIL > but was starting to think someone must have written this better than I > can do quickly I have used LGI for a home-brew solution. Dig up the documentation on LGI$LOGINOUT_CALLOUTS You can preceed and proceed a process (all within the login/out process) with checks for starts, completions, etc. John "REBOOT" Brandon VMS Systems Administrator firstname.lastname.spam.me.not@dalsemi.com