Restarting Job

[Taomyn]

Is there a way to restart a job rather than just retrying it?

I've had a few jobs that failed and a retry won't work but I find that sometimes restarting it by deleting then re-adding the same NZB file does complete. A way to simply "delete and restart" a job would be handy in these cases. I'm using the Glitter theme.

[shypike]

Currently there isn't.
Maybe in a later release.

[Taomyn]

Ok, thanks

[Taomyn]

Any further thoughts on adding this soon? I could have done with it this week with quite a few failed downloads I submitted not being fully available on my provider's servers, but were fine later on.

[ALbino]

Set a propagation delay:

https://sabnzbd.org/wiki/configuration/1.1/switches

[safihre]

Yes, that should solve your issues, shouldn't it?
Just set a delay of about 60 min and surely all articles should be present on your newsserver!

[Taomyn]

Set a propagation delay:

https://sabnzbd.org/wiki/configuration/1.1/switches

Yes, that should solve your issues, shouldn't it?
Just set a delay of about 60 min and surely all articles should be present on your newsserver!

Thanks to both of you it's nice to know that this exists as an option with in sabznbd, but I already have a delay of 60min at the indexer site and this still happens. It's been a balancing act between too low a delay and getting many more of these unready downloads, or too great a delay and the releases being corrupted by DMCA takedowns. Plus it's not just releases not being fully available that causes these failures, simple network issues also, so a simple option to restart in sabznbd would be welcome in my opinion.

[Howitzer]

I have increased my propagation time and that helps but it doesn't really solve the issue. If deleting the previous job and restarting it using the same NZB ALWAYS works (which is my experience on all same-day nzb's) then I would suggest there are two better solutions that are possible:

First choice: An option to "retry download" the job that does more than the current option, it would actually retry to download the missing pieces that might not have fully propagated when the job first ran, it would add to what has already downloaded and complete the job, or it might fail again if the pieces still cannot be found.

Second choice: An option to "restart" which just automates the deletion of the old job and the restart of the process using the same NZB file. While this would still be a huge waste of bandwidth re-downloading articles that were successful on the first try, at least it would provide the capability to fix the failing job using the web/remote interface.

[Taomyn]

Any chance of this making it into the next release v1.2.0?

[shypike]

We could create a scheduler event that will restart failed jobs.
There are limits to retrying. Adding articles to a partial file is not doable, due to a fundamental error in the design of NZB files.

[Taomyn]

We could create a scheduler event that will restart failed jobs.
There are limits to retrying. Adding articles to a partial file is not doable, due to a fundamental error in the design of NZB files.

Wouldn't a restart simply delete all the downloaded files, except the original nzb, then start again? It's not a retry that I would like, but a complete restart of the task.

[shypike]

Normally, already present files are skipped.
You're right, erasing existing ones can be a solution for this case.
We'll see what we can do.

[Taomyn]

Will this make v2? I haven't been able to get v2 installed to see if it's there but I didn't see it mentioned in the change log which was disappointing.

[safihre]

Sorry, don't have the time for it... Could use some coding help for many of the open feature requests :/