Work Unit Results cannot transmit??

QCH

I have 4 WU's waiting on this server... Anyone else? I can ping the server and all my other machines can contact other servers but not this one.

Could not connect to Work Server (results)
[11:38:30] (171.64.122.119:8080)
[11:38:30] - Error: Could not transmit unit 00 (completed February 8). Keeping unit in queue.

t1rhino

Listed as 'busy'. Maybe it needs a reboot.

http://folding.stanford.edu/serverstat.html

QCH

Just found some info on 119 server.... Guess this is happening to a lot of people.

<a href="http://forum.folding-community.org/viewtopic.php?t=7004&quot; target=blank> Folding Community Forum</a>

Which has links to the servers' status... <a href="http://folding.stanford.edu/serverstat.html&quot; Target=blank> Server Status</a>

Straight_Man

QCH2002 wrote:

I have 4 WU's waiting on this server... Anyone else? I can ping the server and all my other machines can contact other servers but not this one.

Could not connect to Work Server (results)
[11:38:30] (171.64.122.119:8080)
[11:38:30] - Error: Could not transmit unit 00 (completed February 8). Keeping unit in queue.

IF you have client 4.0 it will keep retrying every once in awhile, and that client can work with 9 WUs done and one working. If it can connect on a retry, it will send all the WUs that it could not send previously.

Yeah, if a server has alot of WUs out and has sent a lot together it can take a while before it is free before it is neeeding a reboot. Basicly, when it accepts it can send a WU out also. So, when it accepts you can gt another WU from same server. The logic then goes like this-- for each accept, you can have an accept plus a sends of new WU, so an accept can impose a double time load for seerver to keep a connect. Every once in a while folding servers get bunches of return attempts close enough to all at once that they in fact are busy for a time. Every once in while, folding admins have to empty the WUs collected off the servers if tehy have a close to full HD with lots of accepted WUs stored on them. Folding puts emphasis on research, adn makes do with limited storage on servers for acceptance and uses more storage proportionately for WUs that are stuck on server to be sent out, on average than it allocates just for accepting WUs. Part of the reason you see servers maintained so often is to offload data.

Dedicating a box just to collecting WUs sent back costs money and takes handling processing bandwidth on network. Plus, when you dedicate a box to just storage, to handle the feed from namy accepting boxes, it ends up just storing mostly and takes most of resources just to store and then feed the stored stuff to boxes that are used to analyze results.

Over a weekend, Folding is set up good enough that it mostly runs itself except for an ocasssional server outage. When a server dysfunction is temporary or minor, the admins are not called in over the weekend. So you get more of this on weekends. That is why the clients are set up to retry sending and work on. Costs less, more money can go to pure research that way.

John D.

Dragstk

I had this happen about a week ago. It held one WU in queue, but it got another project. Several times it tried to send, finally connected, and sent the waiting WU along. All the time working on another project.
Only happened once, so I took it as a system clitch
Dragstk

Riddick

When this usually happens.. waiting a few hours and resending does the trick most of the time

QCH

I noticed on the server's logs that the server has been "busy" or under load since early sat. morning. Twice durning the weekend, the server was Accepting WU's but only for 20 minutes, or at least it was for some portion between 20 minute updates.