Currently I have a number of backed up transfers due to a server being down. Two from the same project going to a different server uploading fine. However following the,"active" label under the status heading when the first finished the second had (project backoff 05:00:00) after it. This figure reducing as expected with time. The project backoff message appeared despite the fact that the upload continues normally.

I think the 'project backoff' message only clears when the second server upload is complete and acknowledged.

There are some modes of server failure where an upload proceeds to 100% as normal, and only then reports an error: if that happens, the 'project backoff' flag isn't (and arguably shouldn't be) cleared

If I am reading that correctly Richard, you are saying that the bit of the client that gives the message to the manager doesn't know that one project can have more than one upload server? There clearly is not project backoff on the NZ tasks. I have sen this two or three times recently. I don't remember seeing it on my Linux client/manager combinations but then if it only happening when there are two different batches going to different places, I may not have had the right conditions for it at the same time as I was paying attention to what is happening.

Edit: I looked through the event log and there was no record of any problems with the nz uploads going to upload13, only the ones going to the other server.

Yes, we got that fixed in 2011:

- client: problem: suppose a project has 2 data servers,
and one of them is down.
If several downloads from the broken one fail,
we go into "project-level backoff"
and don't start downloads from either server.
(Same applies to uploads).

Solution: make project-level backoff apply only to
transfers that have already failed at least once

Even if uploads are backed off, perhaps because of a broken or inaccessible server, every new upload is tried anyway. If the other server is working and accepts the file, that's it dealt with as normal.

(SVN 24339)