Lost work!

[guru]

I had to change my proxy information at work. I first shut down all the clients, updated the proxy.cfg on all the clients then restarted them. I noticed that all the clients were prompting me to upgrade the client. If I selected no it stopped. If I selected yes it crashed. I then proceeded to manually upgrade all the clients. The new client has different data and therefor does not recognize the old work that has been waiting to be uploaded since yesterday. All that work is lost. At least half a days production which is about 450K units. This auto update function needs to be fixed!

guru

[ColinT]

Yeah, none of my boxes updated correctly, but I did not lose work.

[Angus]

All of mine lost work - I had to delete at least 12 - 18 hours work depending on when I could get to the machines to do a manual update.

Not particularly smooth on their part. I think we can call the first test of the auto-update function a dismal failure.

[IronBits]

I whacked all the WUs sitting in the distribfold directory so I could get on with crunching using the new client...
They should have phased this new thing in... not all at once, cold turkey like.

[Paratima]

I THINK most of my WUs got thru. It's really kinda hard to tell. Rough 24 hours we've had here.

I'm more concerned with wasting horsepower, with the server running so far behind. I run most of my clients on RAMdisk for extra speed, but they're running so fast now, they're also running out of room. :confused: What to do, what to do?

[Paratima]

OK! The 2 AMD Gig-and-a-halfers are going back to Genoming for the duration. Or maybe LGM's.