knreed: "We have worked on a number of items this weekend to both mitigate and further investigate this ongoing issue." in the Known Issues [read only] forum at http://www.worldcommunitygrid.org/forums/wcg/...33481_lastpage,yes#386302.
Applause for your dedication, techs.

Thngs seem to be running much more smoothly from this end. The irregular server outages are hardly noticeable now if at all.

Just a thought - ignore if not relevant.
Some time ago I had an issue with a machine that kept freezing up for periods of 1-2 min and getting WU timeouts and restarts with "exiting with zero status". It turned out to be a dodgy not-very-old system/BOINC data hard drive (WD 500GB Green) that was going into "500GB floppy disc mode", ie slowing right down but never returning even a single hard error, several times per day. The drive passes the WD DOS diagnostics perfectly every time.
Maybe you have a server hardware/firmware problem?

Thanks Kevin for keeping us all in the picture - by the sounds of it, you're making progress, so hopefully, you'll soon have this issue resolved once and for all.

The default on both the switches and the servers was to have both levels of flow-control on. For three years it worked fine. However, the system is behaving much better with TCP flow control only.

However, this is not the root cause - we still see symptoms of the problem appearing, but the tcp-ip flow control is doing a much better job managing through the situation an dramatically reducing the times it leads to an impact to our volunteers. We are continuing to look for the root cause.

Starting to see C4SW uploads stalled again. Anyone else?

Same here on HCC and HFCC.

Cheers