171.67.108.11 & 171.67.108.21 down

[Nick200]

OK, so to get round this inability to upload WUs or get new ones for my old GT240, I have replaced it with a GT750Ti - and it is blazing through new work units for Core 17. It took 10 mins to swap the old one out, worked perfectly after reinstalling the driver and does not seem to be straining my old Dell Inspiron PSU, which was the worry.

That leaves the problem of one old orphan WU for project 5771 sitting in my send queue which, like BertNZ, I will jettison if it does not upload of its own accord in a couple of weeks' time. Pity.

[John_Weatherman]

Sure would be nice to know what's happening with the server and core 11 work. I know it's vacation time, people are away etc etc but work's not being done as I don't know whether to upgrade or not or what to do.

[bruce]

Sure would be nice to know what's happening with the server and core 11 work. I know it's vacation time, people are away etc etc but work's not being done as I don't know whether to upgrade or not or what to do.

I don't know what's happening with the servers, but there's a simple answer: Once PG announces that something will soon be End-Of-Life, there's a good reason to consider an upgrade. Waiting until the last minute isn't really a good idea. Newer generations of hardware use upgraded technology which begins saving power whenever you upgrade. (Of course you have to choose between a low priced GPU with similar PPD and lower power use or a higher priced GPU with a significant increase in PPD.)

[gaitskill]

That’s well and good, but what is it safe to upgrade to.

You can buy the latest and greatest most expensive card and probably be safe. But what if you wanted to buy a somewhat older, maybe used, card? How do you avoid buying a card that is about to become unsupported?

The white list shows what is supported now, but I’m not aware of any list that shows what is near end of life.

It’s been posted that aging cores such as core 11 and core 78 will be retired. Is there a list of cards that are limited to these cores?

Does this server wholly support the cards that are about to become unsupported. If so, is there a list of the cards that it is supporting now?

[bruce]

Welcome to foldingforum.org, gaitskill.

That's a good question, and I'll attempt to provide an equally good answer.

It's really a question of hardware generations and their associated drivers.

Both ATI and NV try to come out with a new generation of GPUs every year or two. Sometimes they succeed and sometimes it takes them longer than they expected. They still need to sell new GPUs, however, and they'd really like you to believe that when they change the left-most digit of the model number, it's a new hardware generation, but they don't actually do that even though that was part of their original plans. Many cards that have an upgraded model number are actually re-branded versions of the previous generation rather than a new model.

When a genuine hardware generation is released, they upgrade the drivers, although it often takes them several driver upgrades to provide full support. Meanwhile, they're still selling the previous generation and in some cases, the generation before that. Drivers still provide excellent support for the previous generation but the quality degrades for earlier generations.

It's not really much different for ATI vs. NV, but I'll use NV as an example.
Core_11 was released when drivers supported what I call the G80 or Tesla generations (though there were a variety of closely related derivatives).
Core_15 was released when drivers supported Fermi.
Core_17 was released when drivers supported Kepler.
Rather than refer to a model number, what I'm calling "generations" can most easily be related to the second letter in the chip name: GMxxx for Maxwell, GKxxx for Kepler, GFxxx for Fermi.

The newest drivers (v 337.88) supposedly support Maxwell, and generally do for most popular games but still have some problems with a couple of essential (to FAH) features. These are the same drivers which have been widely not recommended on this forum because they reduce the speed of chips older than the GK110 which work better if you stay on v327.23 or older.

It should also be noted that when I use "drivers" I'm actually talking about two distinct sets of software. You're responsible for upgrading the drivers on your system. You're welcome to choose any version that provides good support for your hardware. A separate set of driver components is linked into the FahCore. In an ideal world, you can run any working version of the drivers and they're (mostly) compatible with either older or newer versions of the code that's linked into the FahCore -- within reason.

FAH cannot support the G80 hardware forever, simply because NVidia will stop fixing driver problems for that generation of hardware. Considering the fact that this unified shader model was introduced some 7 years ago, it's not surprising.

If you buy a Maxwell (not the same as a GeForce 700 series, bur rather just ones containing the GM107 chip -- or later, once they get them working) you'll get hardware that will be fully supported "soon" and will be so for a "long" time. If you buy a GPU with a GKxxx chip it'll be supported a "long" time, too. Hardware with the GFxxx chip probably won't be supported quite as long. How long depends entirely on NVidia (or ATI/AMD, if that's your preference).

[gaitskill]

Thanks for the explanation. I was wrong thinking that there were specific cards associated each FAH Core. If I understand you right, the association is between a generation of graphic cards/drivers, and the driver (stubs?) of a core.

So, if support for the Tesla generation is discontinued, core 11 is dropped, along with my GTX260.

Thanks also for the welcome. Thought I had registered for the forum when I started folding, 26 June 2009, under the alias of ancient. Had to newly register when I couldn’t find a previous registration.

[farmpuma]

It's been over a week with nothing but crickets, tumbleweed, and rejection.

Seven lonely finished work units with no place to go.

[7im]

Considering core 11 was EOL so long ago, they may just decide this hardware issue is not worth the money or effort to repair if they have sufficient WU returns already, and then call it done.

[John_Weatherman]

Fair enough, but could someone inform us that it's so?

[7im]

Not an unfair request in the least. However, I don't know if they know the answer yet. They may be reviewing the data before either pulling the trigger on the repair, or pulling the plug.

Maybe one of the current Mods that are still in a little better touch with PG than me can prompt someone for an update.

[toTOW]

Yes, we'd like to hear something official ... we hope that it will be a good news, but bad ones have to be announced too if necessary

[toTOW]

Anyone is alive

[VijayPande]

We've been working to back up that machine and get all of the data off. I think the RAID is about to go and we'll retire that server. With it, that probably means the retirement of core11.

Some background on how we handle old projects: old projects are often very useful to keep running at a slow drip, back burner style since it can lead to a few very long MD trajectories (say over 2 to 3 years). That's very much been the case for this server. We've been analyzing the data set we have now compared to 2010 (this is the Voelz et al paper on NTL9) and it's pretty amazing to see what 4 years has brought us.

One might ask then maybe we should just start projects like this. However, it's probably not worth the effort to start up projects like this since we've got other things going on but leaving these on the back burner and seeing what happens is also reasonable when that's easy for us to do (a lot of this issue is the man power to run all of these projects). And, for old cores, it doesn't make sense to start up new projects with them since they are so limited compared to newer GPU cores.

[barebear]

Hi Bruce, Jesse V , 7IM, bollix 47 (hi Mike), panther x and all the other people that have helped in the past.....

I'm running 64 bit Win 7 Professional with a core I7 860 2.8 GHz and Nvidia GTS 250 and 16 Gb ram.

The cpu shows as slot 00 cpu:7 -- this cpu has 8 cores and is getting and doing work assignments w/ no issues that I can detect other than its description in the Advanced Control main window.

The GPU can't get work assignments -- the log is here:

Code: Select all

*********************** Log Started 2014-07-31T05:10:56Z ***********************
05:10:56:************************* Folding@home Client *************************
05:10:56:      Website: http://folding.stanford.edu/
05:10:56:    Copyright: (c) 2009-2014 Stanford University
05:10:56:       Author: Joseph Coffland <[email protected]>
05:10:56:         Args: --open-web-control
05:10:56:       Config: C:/Users/Peter/AppData/Roaming/FAHClient/config.xml
05:10:56:******************************** Build ********************************
05:10:56:      Version: 7.4.4
05:10:56:         Date: Mar 4 2014
05:10:56:         Time: 20:26:54
05:10:56:      SVN Rev: 4130
05:10:56:       Branch: fah/trunk/client
05:10:56:     Compiler: Intel(R) C++ MSVC 1500 mode 1200
05:10:56:      Options: /TP /nologo /EHa /Qdiag-disable:4297,4103,1786,279 /Ox -arch:SSE
05:10:56:               /QaxSSE2,SSE3,SSSE3,SSE4.1,SSE4.2 /Qopenmp /Qrestrict /MT /Qmkl
05:10:56:     Platform: win32 XP
05:10:56:         Bits: 32
05:10:56:         Mode: Release
05:10:56:******************************* System ********************************
05:10:56:          CPU: Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz
05:10:56:       CPU ID: GenuineIntel Family 6 Model 30 Stepping 5
05:10:56:         CPUs: 8
05:10:56:       Memory: 16.00GiB
05:10:56:  Free Memory: 13.64GiB
05:10:56:      Threads: WINDOWS_THREADS
05:10:56:   OS Version: 6.1
05:10:56:  Has Battery: true
05:10:56:   On Battery: false
05:10:56:   UTC Offset: -7
05:10:56:          PID: 5764
05:10:56:          CWD: C:/Users/Peter/AppData/Roaming/FAHClient
05:10:56:           OS: Windows 7 Professional
05:10:56:      OS Arch: AMD64
05:10:56:         GPUs: 1
05:10:56:        GPU 0: NVIDIA:1 G92 [GeForce GTS 250]
05:10:56:         CUDA: 1.1
05:10:56:  CUDA Driver: 6000
05:10:56:Win32 Service: false
05:10:56:***********************************************************************
05:10:56:<config>
05:10:56:  <!-- Network -->
05:10:56:  <proxy v=':8080'/>
05:10:56:
05:10:56:  <!-- Slot Control -->
05:10:56:  <power v='FULL'/>
05:10:56:
05:10:56:  <!-- User Information -->
05:10:56:  <passkey v='********************************'/>
05:10:56:  <team v='39340'/>
05:10:56:  <user v='Peter_Baerwald'/>
05:10:56:
05:10:56:  <!-- Folding Slots -->
05:10:56:  <slot id='0' type='CPU'/>
05:10:56:  <slot id='1' type='GPU'/>
05:10:56:</config>
05:10:56:Trying to access database...
05:10:56:Successfully acquired database lock
05:10:56:Enabled folding slot 00: READY cpu:7
05:10:56:Enabled folding slot 01: READY gpu:0:G92 [GeForce GTS 250]
05:10:56:WU02:FS00:Starting
05:10:56:WU02:FS00:Running FahCore: "C:\Program Files (x86)\FAHClient/FAHCoreWrapper.exe" C:/Users/Peter/AppData/Roaming/FAHClient/cores/web.stanford.edu/~pande/Win32/AMD64/Core_a3.fah/FahCore_a3.exe -dir 02 -suffix 01 -version 704 -lifeline 5764 -checkpoint 15 -np 7
05:10:56:WU02:FS00:Started FahCore on PID 4660
05:11:00:WU02:FS00:Core PID:5844
05:11:00:WU02:FS00:FahCore 0xa3 started
05:11:00:WU02:FS00:0xa3:
05:11:00:WU02:FS00:0xa3:*------------------------------*
05:11:00:WU02:FS00:0xa3:Folding@Home Gromacs SMP Core
05:11:00:WU02:FS00:0xa3:Version 2.27 (Dec. 15, 2010)
05:11:00:WU02:FS00:0xa3:
05:11:00:WU02:FS00:0xa3:Preparing to commence simulation
05:11:00:WU02:FS00:0xa3:- Looking at optimizations...
05:11:00:WU02:FS00:0xa3:- Files status OK
05:11:00:WU00:FS01:Connecting to 171.67.108.201:80
05:11:00:WU02:FS00:0xa3:- Expanded 3810693 -> 4136808 (decompressed 108.5 percent)
05:11:00:WU02:FS00:0xa3:Called DecompressByteArray: compressed_data_size=3810693 data_size=4136808, decompressed_data_size=4136808 diff=0
05:11:00:WU02:FS00:0xa3:- Digital signature verified
05:11:00:WU02:FS00:0xa3:
05:11:00:WU02:FS00:0xa3:Project: 6098 (Run 6, Clone 29, Gen 493)
05:11:00:WU02:FS00:0xa3:
05:11:00:WU02:FS00:0xa3:Assembly optimizations on if available.
05:11:00:WU02:FS00:0xa3:Entering M.D.
05:11:06:WU02:FS00:0xa3:Using Gromacs checkpoints
05:11:06:WU02:FS00:0xa3:Mapping NT from 7 to 7 
05:11:07:WU02:FS00:0xa3:Resuming from checkpoint
05:11:07:WU02:FS00:0xa3:Verified 02/wudata_01.log
05:11:07:WU02:FS00:0xa3:Verified 02/wudata_01.trr
05:11:07:WU02:FS00:0xa3:Verified 02/wudata_01.edr
05:11:08:WU02:FS00:0xa3:Completed 257395 out of 500000 steps  (51%)
05:11:21:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Failed to connect to 171.67.108.201:80: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond.
05:11:21:WU00:FS01:Connecting to 171.64.65.160:80
05:11:42:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond.
05:11:42:ERROR:WU00:FS01:Exception: Could not get an assignment
05:11:42:WU00:FS01:Connecting to 171.67.108.201:80
05:11:42:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Empty work server assignment
05:11:42:WU00:FS01:Connecting to 171.64.65.160:80
05:11:44:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: No connection could be made because the target machine actively refused it.
05:11:44:ERROR:WU00:FS01:Exception: Could not get an assignment
05:12:42:WU00:FS01:Connecting to 171.67.108.201:80
05:12:43:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Empty work server assignment
05:12:43:WU00:FS01:Connecting to 171.64.65.160:80
05:12:44:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: No connection could be made because the target machine actively refused it.
05:12:44:ERROR:WU00:FS01:Exception: Could not get an assignment
05:14:20:WU00:FS01:Connecting to 171.67.108.201:80
05:14:20:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Empty work server assignment
05:14:20:WU00:FS01:Connecting to 171.64.65.160:80
05:14:21:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: No connection could be made because the target machine actively refused it.
05:14:21:ERROR:WU00:FS01:Exception: Could not get an assignment
05:16:57:WU00:FS01:Connecting to 171.67.108.201:80
05:16:57:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Empty work server assignment
05:16:57:WU00:FS01:Connecting to 171.64.65.160:80
05:16:58:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: No connection could be made because the target machine actively refused it.
05:16:58:ERROR:WU00:FS01:Exception: Could not get an assignment
05:18:41:WU02:FS00:0xa3:Completed 260000 out of 500000 steps  (52%)
05:21:11:WU00:FS01:Connecting to 171.67.108.201:80
05:21:11:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Empty work server assignment
05:21:11:WU00:FS01:Connecting to 171.64.65.160:80
05:21:12:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: No connection could be made because the target machine actively refused it.
05:21:12:ERROR:WU00:FS01:Exception: Could not get an assignment
05:28:02:WU00:FS01:Connecting to 171.67.108.201:80
05:28:02:WARNING:WU00:FS01:Failed to get assignment from '171.67.108.201:80': Empty work server assignment
05:28:02:WU00:FS01:Connecting to 171.64.65.160:80
05:28:04:WARNING:WU00:FS01:Failed to get assignment from '171.64.65.160:80': Failed to connect to 171.64.65.160:80: No connection could be made because the target machine actively refused it.
05:28:04:ERROR:WU00:FS01:Exception: Could not get an assignment
05:33:23:WU02:FS00:0xa3:Completed 265000 out of 500000 steps  (53%)

Please advise what might be the problem and how to fix it.

Mod edit - Added code tags to log

[bollix47]

Hi barebear (Peter) ... as you can see I've moved your post to the thread discussing the problem. The server that produces work for your GTS 250 is not in service at this time and from Dr. Pande's post it appears that core 11 work may have come to an end. That probable eventuality was mentioned in the blog quite some time ago. For now there's nothing you can do to 'fix' the problem other than upgrading your GPU to a more recent model (reading through this thread may give you some ideas if that's the route you decide to take). You can remove or pause your GPU slot so that the client doesn't waste time trying to get somewhere that is now unreachable.