171.64.65.56 not responding

[kasson]

Thanks for the heads up. I think we got the problem; the server is coming back up right now.

[314159]

Once again, I (and we in the affected community) thank you.

The time lag in getting the server back up was certainly within acceptable parameters regardless of my earlier rant.

I do, however, contend that some of my feelings are shared by others in the OSX/Linux "volunteer community".

They also share my respect for you, Dr. Kasson, for your research most importantly but also for the "many other things" that you do here in the Forum and elsewhere.

We do, of course, acknowledge the magnitude of "those other platforms" and cannot really criticize present PG priorities.

That said:

Do I detect a sense of humor in all of this since the last two machines to complete received new work from "good old 171.64.65.56"?

[susato]

Thanks Peter! My mini connected successfully around 18 minutes ago to re-download a unit trashed by a routine software update & restart. Everything looks "nominal" with the server now.

[spazzcat]

I am still getting 400 errors on this server. At this point I see no reason to keep wasting server bandwidth on folding...

[codysluder]

I am still getting 400 errors on this server. At this point I see no reason to keep wasting server bandwidth on folding...

Error 400 = Bad Request

You may have already tried this, but I have to suggest it in case you didn't:
Stop the client / delete the FahCore / reboot / restart the client.

Since others seem to be getting through, there's a good chance the problem is at your end.

[kasson]

The server was in reject this morning; we restarted the binary.

[314159]

Rejecting once again plus the "alleged Collection Server" remains generally impotent.

Anyone around this weekend to restart the binary?

Please Help!!!

Thanks!

John

[codysluder]

It looks like the binary has been restarted.

The only hope for the "alleged Collection Server" is faster hardware plus the new server code. Each CS does accept a continuous stream of uploads, but they're designed to handle temporary overloads of servers that normally handle their entire workload, and most of the WorkServers have not been able to keep up with their normal workload for some time, now.

Of course there's at least one positive way of looking at it: FAH has more donors than they planned on.

[314159]

Hey codysluder,

Yup. It was restarted shortly after I posted.
It's good to know that the PG has someone available to keep things running on weekends.
It's a shame that the PG has to work on weekends!
We all do appreciate them, of course.
So thanks to whomever reset this server (or thanks to the code that did it "auto-magically")

I agree about the need for a fix on that damn C.S. It has become a true thorn in my side.
Note that I was not even getting a "503" and could not ping the C.S. at the time, so I suspect that it has absolutely nothing to do with the number of donors in this case. (note the Wservers that it "allegedly" backs)
The load on the C.S. was within acceptable parameters at the time and one WU did in fact get sent early on.
(I am curious as to whether it will ever show up in our Team Stats).

I am not particularly over-joyed with the new WUs that I received but they will get done and the science will hopefully benefit.

I hope that our PG friends are able to enjoy the balance of their much deserved "free time".

[314159]

Um, (13-July-09 0300 - 04:56 PDT)

This server is once again in REJECT status and the collection server is dishing out its typical "503's"..

May I request a restart of this server once one of the kind folks at the Pande Group wake up?

Thank you!

[kasson]

Done...sorry for the delay.

[314159]

Um,

I assure you that we ALL assume that you should get a good night's sleep, and often wonder whether you do.

We appreciate your prompt (vs. delayed) assistance.

BTW, it gave me the opportunity to experiment a bit with the CS behind this WS.
Employing an appropriate delay between -send xx's from several machines, about 1/4 were successfully sent and acknowledged on the first or second attempt.
The others either received 503's or more typically were unable to connect until 5 or 6 tries.

I guess that this is better than nothing or wasting bandwidth (as in the past) only to find that the CS had no record of the WU.
I would still question whether this CS is "robust" enough to serve its intended purpose.

All completed WUs are now sitting happily on the CS at this time (or were) - plus I had an opportunity to clear out the work directory detritus (that the client/core should have done) from the machines involved.

THANK YOU for your continuing diligence and support;- - - above and beyond typical expectations!

[Ravage7779]

It's rejecting again. Going to be interesting to see how long a p5102 runs...

[314159]

REJECT STATUS.....543PM PDT 26 July.

[kasson]

Got it--thanks