Did I just learn a difficult lesson with usernames/passkeys?

[nshorts_nhs]

Hello all,

My team switched our active nodes to the same username and passcode for all units. I now have one (low-spec, 6 core) server stuck on a 14.69 day project.

Newbie questions:
1- Does the passkey need to match the username? My assumption was no.
2- Is there a definitive guide to dumping this workload under 7.5.1 (Linux)? The forums and google have not shown any commands that worked.

Thank you!

[davidcoton]

1 The passkey is locked to the email address, IIRC. I believe you can change username keeping the passkey (but have not tried myself).
2 Dumping is not generally available and is strongly discouraged. Please post your log (from Advanced Control, Refresh before Copy) in Code tags here so someone can check what is happening.

[Neil-B]

There have been a small number of malformed WUs which are taking "FAH" too long (sorry couldn't resist it ... The log will let this be checked and if it is it can be reported (the were then dumped with approval/blessing from the team)

[Joe_H]

Same username and email address on the application form gets you sent the same passkey. Once you have a passkey, you can use it with as many usernames as you want, does not have to match the one used to get the passkey. Only limitations is that each username/passkey pair needs to work on and return 10 WUs before becoming eligible for the bonus.

[nshorts_nhs]

1 The passkey is locked to the email address, IIRC. I believe you can change username keeping the passkey (but have not tried myself).
2 Dumping is not generally available and is strongly discouraged. Please post your log (from Advanced Control, Refresh before Copy) in Code tags here so someone can check what is happening.

It rolled, but I'll grab the older one when I get a chance:

Code: Select all

14:38:39:Switching to user fahclient
14:38:39:Trying to access database...
14:38:39:Successfully acquired database lock
14:38:39:Enabled folding slot 00: READY cpu:6
14:38:39:WU01:FS00:Starting
14:38:39:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/cores.foldingathome.org/v7/lin/64bit/avx/Core_a7.fah/FahCore_a7 -dir 01 -suffix 01 -version 705 -lifeline 517 -checkpoint 15 -np 6
14:38:39:WU01:FS00:Started FahCore on PID 526
14:38:39:WU01:FS00:Core PID:530
14:38:39:WU01:FS00:FahCore 0xa7 started
14:38:40:WU01:FS00:0xa7:*********************** Log Started 2020-03-30T14:38:39Z ***********************
14:38:40:WU01:FS00:0xa7:************************** Gromacs Folding@home Core ***************************
14:38:40:WU01:FS00:0xa7:       Type: 0xa7
14:38:40:WU01:FS00:0xa7:       Core: Gromacs
14:38:40:WU01:FS00:0xa7:       Args: -dir 01 -suffix 01 -version 705 -lifeline 526 -checkpoint 15 -np 6
14:38:40:WU01:FS00:0xa7:************************************ CBang *************************************
14:38:40:WU01:FS00:0xa7:       Date: Nov 5 2019
14:38:40:WU01:FS00:0xa7:       Time: 06:06:57
14:38:40:WU01:FS00:0xa7:   Revision: 46c96f1aa8419571d83f3e63f9c99a0d602f6da9
14:38:40:WU01:FS00:0xa7:     Branch: master
14:38:40:WU01:FS00:0xa7:   Compiler: GNU 8.3.0
14:38:40:WU01:FS00:0xa7:    Options: -std=c++11 -O3 -funroll-loops -fno-pie -fPIC
14:38:40:WU01:FS00:0xa7:   Platform: linux2 4.19.0-5-amd64
14:38:40:WU01:FS00:0xa7:       Bits: 64
14:38:40:WU01:FS00:0xa7:       Mode: Release
14:38:40:WU01:FS00:0xa7:************************************ System ************************************
14:38:40:WU01:FS00:0xa7:        CPU: Intel(R) Xeon(R) CPU E5-2603 v3 @ 1.60GHz
14:38:40:WU01:FS00:0xa7:     CPU ID: GenuineIntel Family 6 Model 63 Stepping 2
14:38:40:WU01:FS00:0xa7:       CPUs: 6
14:38:40:WU01:FS00:0xa7:     Memory: 46.88GiB
14:38:40:WU01:FS00:0xa7:Free Memory: 42.28GiB
14:38:40:WU01:FS00:0xa7:    Threads: POSIX_THREADS
14:38:40:WU01:FS00:0xa7: OS Version: 3.10
14:38:40:WU01:FS00:0xa7:Has Battery: false
14:38:40:WU01:FS00:0xa7: On Battery: false
14:38:40:WU01:FS00:0xa7: UTC Offset: -5
14:38:40:WU01:FS00:0xa7:        PID: 530
14:38:40:WU01:FS00:0xa7:        CWD: /var/lib/fahclient/work
14:38:40:WU01:FS00:0xa7:******************************** Build - libFAH ********************************
14:38:40:WU01:FS00:0xa7:    Version: 0.0.18
14:38:40:WU01:FS00:0xa7:     Author: Joseph Coffland <joseph@cauldrondevelopment.com>
14:38:40:WU01:FS00:0xa7:  Copyright: 2019 foldingathome.org
14:38:40:WU01:FS00:0xa7:   Homepage: https://foldingathome.org/
14:38:40:WU01:FS00:0xa7:       Date: Nov 5 2019
14:38:40:WU01:FS00:0xa7:       Time: 06:13:26
14:38:40:WU01:FS00:0xa7:   Revision: 490c9aa2957b725af319379424d5c5cb36efb656
14:38:40:WU01:FS00:0xa7:     Branch: master
14:38:40:WU01:FS00:0xa7:   Compiler: GNU 8.3.0
14:38:40:WU01:FS00:0xa7:    Options: -std=c++11 -O3 -funroll-loops -fno-pie
14:38:40:WU01:FS00:0xa7:   Platform: linux2 4.19.0-5-amd64
14:38:40:WU01:FS00:0xa7:       Bits: 64
14:38:40:WU01:FS00:0xa7:       Mode: Release
14:38:40:WU01:FS00:0xa7:************************************ Build *************************************
14:38:40:WU01:FS00:0xa7:       SIMD: avx_256
14:38:40:WU01:FS00:0xa7:********************************************************************************
14:38:40:WU01:FS00:0xa7:Project: 13821 (Run 318, Clone 2, Gen 83)
14:38:40:WU01:FS00:0xa7:Unit: 0x0000005f80fccb095c88395fa7250b60
14:38:40:WU01:FS00:0xa7:Digital signatures verified
14:38:40:WU01:FS00:0xa7:Calling: mdrun -s frame83.tpr -o frame83.trr -x frame83.xtc -cpi state.cpt -cpt 15 -nt 6
14:38:40:WU01:FS00:0xa7:Steps: first=10375000 total=10375000
14:38:45:WU01:FS00:0xa7:Completed 1671702 out of 10375000 steps (16%)
18:27:01:WU01:FS00:0xa7:Completed 1763750 out of 10375000 steps (17%)
******************************* Date: 2020-03-30 *******************************
22:44:03:WU01:FS00:0xa7:Completed 1867500 out of 10375000 steps (18%)
03:00:36:WU01:FS00:0xa7:Completed 1971250 out of 10375000 steps (19%)
******************************* Date: 2020-03-31 *******************************
07:16:56:WU01:FS00:0xa7:Completed 2075000 out of 10375000 steps (20%)
11:34:00:WU01:FS00:0xa7:Completed 2178750 out of 10375000 steps (21%)

There have been a small number of malformed WUs which are taking "FAH" too long (sorry couldn't resist it ... The log will let this be checked and if it is it can be reported (the were then dumped with approval/blessing from the team)

Good one Done.

Same username and email address on the application form gets you sent the same passkey. Once you have a passkey, you can use it with as many usernames as you want, does not have to match the one used to get the passkey. Only limitations is that each username/passkey pair needs to work on and return 10 WUs before becoming eligible for the bonus.

Not so concerned about bonus, just wondering if this anomaly was the result of having multiple hosts with the same name/passkey- they all shared the same passkey with different user/node names before. Possibly the assignment servers thought the resources were all-in-one or similar, which in hindsight... doesnt make much sense, but could still be a remote possibility. Right now the ETA is 14.02 days, expiration on 2020-04-03.

Modsizeit: Added Code Tags - PantherX

[ipkh]

From other posts it seems you got a badly formed project. Apparently the clu projects should have about 125,000 steps at the most. It is fixed fir future work generation but a few fell through the cracks and couldn't be pulled.

[Neil-B]

When Joe_H (I tend to defer to his knowledge on this issue - Sorry Joe) sees this he'll confirm but the 10m+ steps I think means malformed - he can check if it is one of the known cases … If it would complete before expiry the science would actually still be good - but since it won't I am fairly sure he will confirm best way to dump it.

[davidcoton]

Yes, check that the expected completion is beyond the expiry, if so dump it. It's WU01 in the log, so pause that, go to /var/lib/fahclient/work, and delete the directory 01.
I've notified the Project Manager.

[nshorts_nhs]

Yes, check that the expected completion is beyond the expiry, if so dump it. It's WU01 in the log, so pause that, go to /var/lib/fahclient/work, and delete the directory 01.
I've notified the Project Manager.

Done, thanks! Didn't want to do anything without bringing it up on the forums first.


Other question (to all for confirmation)- so same name across multiple nodes shouldn't cause an issue?

[uyaem]

Other question (to all for confirmation)- so same name across multiple nodes shouldn't cause an issue?

I fold on two machines using the same username and key, no issues.
Everything works as expected - from receiving and transmitting work packages to point crediting.

[nshorts_nhs]

Other question (to all for confirmation)- so same name across multiple nodes shouldn't cause an issue?

I fold on two machines using the same username and key, no issues.
Everything works as expected - from receiving and transmitting work packages to point crediting.

Thank you

[Joe_H]

Yes, at over 10 Million steps this is a bad WU, go ahead and dump.

From other posts it seems you got a badly formed project. Apparently the clu projects should have about 125,000 steps at the most. It is fixed fir future work generation but a few fell through the cracks and couldn't be pulled.

This particular project, and some related to it should have 125,000 total steps. But there are other CPU projects where the normal number could be 250K, 500K, or more. Only quick way to tell is if you have a log entry for that project that you can compare with.

I have an extensive collection of past logs to search through, but many of these recently added projects have not shown up in them yet. I can check with people who have tested these WUs before release though, and follow up here on the forum on reports like this.