[Rock-ons] Address timeout issues during Rock-on list refresh cycle

[Hooverdan96]

Thanks to greven and jimla1965 for reporting an issue that seems to intermittently occur during the Rock-on repository update and processing on a Rockstor appliance, leading to timeout issues (the first instance reported was intermittent, the second instance reported was persistent).

https://forum.rockstor.com/t/another-take-on-unknown-internal-error-doing-a-post-to-api-rockons-update/7739/

and, more recently,

https://forum.rockstor.com/t/unknown-internal-error-doing-a-post-to-api-rockons-update-old-fix-gone/9035/

The current symptom can be observed in the Rockstor WebUI.

Unknown internal error doing a POST to /api/rockons/update

Further details can be gleaned from the two posts listed above.

[Hooverdan96]

For reference, @phillxnet is working on a promising approach to substantially shorten the retrieval time (and possibly avoid above errors related to timeouts): #2707 (with related PR #2708

[phillxnet]

@Hooverdan96 Lets see if we can get some field confirmation on the linked issue/pull-request once we get the next testing channel rpm out. If we have favourable feedback I propose that we close this one against PR #2708, and potentially also #2706 as I think they will hopefully both help out on this front.

The next testing channel rpm version to contain both hopefully related fixes should be 5.0.5-0

[phillxnet]

@Hooverdan96 Maybe we can close this one now - we have not explicitly had confirmation but we have significantly reduced this whole process, and made it far less prone to time-out, i.e. via single https session use, and more appropriate gunicorn threads.

[Hooverdan96]

Yes, I think for now we can close it, based on your assessment and all the recent changes that have been merged into the latest testing channel. If during the real-world application it creeps up again, we'll create a new issue.