How to queue nightly backups with networker 6.1.1

[RobinPiepjohn]

Hi

Iwant to queue or control backups in the way that a group may be startet but only if no other jobs are running.
If there is another then I want waiting until ther is no job running.
In my configuration I start a group at 00:00 that is running 1:00h
At 01:00 there startet a backup Job from a SAP on Oracle client.
If the midnight job runs longer than 1 hour I have conflicts backing up both jobs.Sometimes the system cannot backup some oracle logs, sometimes the nidnight group stops by timeaouts

Which posibilities are available to control all backups so I didn't have these conflicts.
Note:After Sap BAckup there are starting more groups until the early morning

Please tell me out of your expirience how to do

Many thanks

 

[TDun]

You can set your parallelism so that only the requisite number of streams can run.
To get EXACTLY what you seem to be suggesting, you weould need to make sure that you only generate one saveset per backup or that your parallelism is down to 1 (which would be extremely inefficient). Or am I misunderstanding you?

 

[mkkeane]

If your server is UNIX, then you can start and stop groups from command line using savegrp commands. If this is an NT server, then scripting is a little more tricky but since legato is a command line heavy application, with much effort and skill, a scheduler (third party) could be used to manage the jobs.

Hope this leads to a solution for you.

Matthew

 

[BW78]

You can start both backup jobs with using one Batch-script. start this script with a scheduler like winat(NT) or at (nt command line) on the Backup server.

---> winat (every day Start 00:00 midnight.cmd)

-- midnight.cmd --
savegrp SAPBackup (-w -y -l -c)
savegrp next backup group (-w -y -l -c)

---------------------

this is a very easy script, perhaps it helps.

 

[BW78]

If u want to start a group only when no other Backup group is startet, then u have to ckeck the Processes

if the savegrp.exe is running then the next backup job have to wait.

after the midnight-group wich starts at 00:00
you start the next group with a Batch-script.

In this batch script u have to check the savegrp.exe
In windows NT i can get a Process list with the tlist command

start
Tlist |findstr savegrp.exe
if %errorlevel%==0 then wait 5
goto start
savegrp (next backup group)

 

[jkalbus]

The only problem with this kind of waiting loop is, that the networker backup ends with timeout errors, if the loop takes too long to finish, i.e. the backup job that has to be finished.
Tried that in Unix - which is almost the same and got every now and then timeout errors. Increasing timeout to about 5 to 6 hours helped - but for me that is only a temporary help.

Until now we are still looking for a good scheduler to help us solve some problems, but as it looks you have to use a third party scheduler to integrate even external processes.

Johanes

 

[BW78]

Why you don´t set the timeout time of the Client to "0"?
Then the CLient will never timout.!