Jobs schedule best practices

[JosueM]

Good day everyone,

Is there any recommendation or best practices for job scheduling. I used to group similar servers per job then I set one job to start lets say at 9:00pm and the other jobs chaining one after the other. Initially I tough this was a good way to plan the schedule but it seems to have some drawbacks like I run manually a job then all the job following the chain will also start after. currently I'm using v7 and planning to upgrade to v8 in the following weeks.

Thanks for any help on this.

[PTide]

Hi,

Actually, chaining is not considered to be a good practive. Please take a look at this post.

Could you please describe your backup policy and goals?

Thank you.

[foggy]

Right, we typically do not recommend jobs chaining due to undesired consequences it incurs (like the one you've mentioned). Just enable parallel processing, configure the maximum number of concurrent tasks allowed on proxy/repository, start the jobs one by one with a short interval (say, 5 minutes), and let automatic resource scheduler do its job.

[JosueM]

Thanks for your reply.

Ptide: currently we backup locally from production NAS to another NAS using hotadd or network and reverse incremental to save space, then we have backup copy jobs to a remote site for critical vms. Also we do replicate critical server to remote site. this happens once per day. Our goal is protect the entire worload local and remote and if possible to have some backups to run more than once per day during production hours but is pretty complicated beacuase it hits pretty hard the performance when the backup runs especially when create/remove the snapshots.

Thanks.

[PTide]

some backups to run more than once per day during production hours

Out of curiosity - what kind of servers do you need to backup during production hours?

[JosueM]

they are application servers, they usually run enterprise resource planning, sales and HR applications and MS-SQL as data repository, there is also a exchange server.

[PTide]

You can reduce the performance hit by using transaction logs backup. That would allow you to have a "point-in-time recovery" restore point for you SQL server without taking a snapshot each time. Please refer to this article to see how it works.

[JosueM]

PTide thanks for your help. I'll give a try to this way to backup a sql server.