Panic Mode On (78) Server Problems?

Mine are being delivered, but never show up in the listing once uploaded. This has been happening for some time now. For me its not about the numbers or credit, it is about if the CPU time is serving a bigger picture. Not sure if there is an issue that needs to be looked at if results are being dropped and not entered into the database.
____________

The astropulse tapes are almost splitted. Only 25 channels remain. maybe it will help
____________

Maybe a suggestion to those windows users having download problems you can try this, it helped for me.

*DO THIS AT YOUR OWN RISK*

For me Windows 7 was not setting TCP window size/TCP optimizations correctly. Since my last reload I had not run this and though I would try it. It corrected most of my network connection problems. This is not perfect but I am no longer getting TCP connection issues and downloads/uploads are proceeding more smoothly. Program must be run as administrator, just select the Optimal radio button at bottom and apply. Then reboot.

SG TCP Optimizer info about program:
http://www.speedguide.net/tcpoptimizer.php

Download here:
http://www.speedguide.net/downloads.php

So far have downloaded 100+ tasks with only minor retry's on some of the downloads. Uploads are going though just about every time on the first shot, they were piling up. And this is with NNT off. Was tired of the constantly allowing/disallowing tasks, seems to be going reasonably fine now.

So for myself this looks like mostly a windows networking issue. For others it may be a differnt problem. I would only recomend using the Optimal radio button in the program unless you really know what your doing and fully understand how the TCP protocol works.

If anyone else tries this report back if it works for you. Or if it doesn't.

*DO THIS AT YOUR OWN RISK*

This is just my 0.02 cents.

____________

It seems to me (I am not a code person) that the problem back at the project level is pretty clearly with the scheduler.

Typically, when there is no work to offer, or no work to offer for a given client, when the scheduler is working properly, it sends back a reply -- no work available, and the client then moves on to other tasks.

The current messed up SETI scheduler is responding to new work requests with 'looking', 'looking' -- and is hung up in a loop.

Only when the client (after 5 minutes or so) realizes that the scheduler is definitely out to lunch, does the client then move on to other tasks.

This problem AT THE SERVER creates a two part problem, -- at the client side, it impedes communications with other projects (or reporting of work to SETI) as the 'phone is off the hook'. At the project side, it creates a major traffic problem as client scheduler connections, which typically run a minute or two are all running 5 minutes or longer.

OK -- that much seems pretty obvious at a layman level.

For users (we are all users here), I think the current frustration has several components.

First -- I've really not seen anything from the project that suggests they have looked at this problem yet -- I suspect they have, but the information vacuum is troublesome. I realize Matt isn't around at the moment, so our primary project communications 'good guy' isn't in the loop.

Second -- this problem has persisted now for 10 days or more -- in a project of this size that is quite severe.

Third -- folks have run out of work to process -- and that adds to the level of angst folks have.

At this point it seems the best approach is for everyone to go to 'no new work' mode and for somebody from the project to communicate.

For users (we are all users here),

I was sort of wondering a bit -- since corporations are people, are CPU's people, and if CPU's are people, are programs people? <grin>

cdemers wrote
"For me Windows 7 was not setting TCP window size/TCP optimizations correctly. Since my last reload I had not run this and though I would try it. It corrected most of my network connection problems"

Hmm my Windows 7 fix for the "Lost / Ghost Tasks clogging up the scheduler"
was to use a Proxy / Concentrator I thought at first this was due to the Proximity of the Proxy server (San Jose) to SETI, but it is much more likely
now that cdemers has seen good results with TCP changes that the Proxy server
does not use Windows.

Good work cdemers now does this help anyone else or just me?

PS - I could not get any downloads to work until I started using a proxy
I have not had any problems scheduling work or downloading or uploading since
06=Nov when I switched to a proxy server
____________

Well I have put this machine on a http proxy server and going like a bat out of hell except every task is a GHOST task
All resends, every one.

Here is the proxy address I am using
201.147.20.245:80

Not sure where it is from but it works.

Now if I just get by the limits

Michael Miles
The Assimilators

The proxy is working. Downloads are very slow..but sure beats getting nothing.

Can someone suggest one that might be faster?

Tom, I do notice in the Tools option in the Boinc Management Menu there are ways of setting connections with two kinds of proxies: One is a socks proxy and there is one other option. Is this where I would put my proxy internet ID numbers?

The next question is how to find these proxies. I am not at all sure how to do that and where to look to find it. I use Comcast as my ISP and have an enhanced help service available to me.

I am running pretty well with my notebook computers except for the very powerful one which is now not able to download much at all when I set to no new tasks. My two desktops with good Nvidia graphics cards are gradually running out of work.

Perhaps you or some others here that are informed and good explainers of setting up the proxies can give me some informed finding aids and tools here. I am willing to work at it once I know how to start. Thanks in advance for any help you can give me. It will probably help many others too. I am sorry to say this, but we are really getting very little help out of the project as this grinds many of us to a halt. Brother Frank

Just as a followup, everything still running much more smoothly. Only getting the odd download that needs a retry. Uploads have been going ok, just a little slow. Downloads have picked up speed.

____________

My RAC has been steadily declining and I have noticed that the task list for two of my rigs show most of the assigned tasks under Error, and the status as abandoned. Could anyone tell me why this might occur. The rigs still have all the tasks and are crunching them, but obviously, not gaining any credit for the work being done. Should I reset the rig or is this something that will get sorted out automatically?