Page 1 of 2 12 LastLast
Results 1 to 40 of 53

Thread: NEEDED: people who can make the current linux client crash

  1. #1

    NEEDED: people who can make the current linux client crash

    I know a lot of people can only make 1 or 2 versions of the linux client work. For some, v1.0 works and v1.02 crashes. For others, v1.0 crashes and v1.02 works. For some, v1.0 and v1.02 both crash and only v0.97 works.

    If I release a new version of the SB client for linux, I'd like to only have to release one.

    Anyone who runs the linux client, this is what I want from you:

    1) A list of each version of the client that runs and crashes on your system(s) between version v0.97 and v1.02. I don't need to know about network problems as those are fixed in the v1.10 code, I just need to know about seg faulting. Something like:

    RedHat 8 box (kernal 2.xx.xx)
    v0.97 -- runs, never crashes
    v1.00 -- crashes immediately on startup
    v1.01 -- runs, no crashes
    v1.02 -- runs, crashes on submitting sometimes

    give me info for as many different boxes as you can and as many versions of the client as you can. if you need diff versions for testing, they are all still available here: http://linux.redbird.com/~alien88/ don't bother testing below v0.97 as those clients can no longer connect to the server. there are several versions but at least test v0.97, v1.00, and v1.02. i recommend changing your user to 'supersecret' during testing, that way you can finish tests in a few minutes to see if they will crash after a submission.

    2) Your email address if you want to be involved in testing the next version before it comes out.

    When I get enough testers with enough different problems, I'll build some new versions until I get one that works for everyone. If we get enough people to participate, this should help create a new linux/FreeBSD/BeOS client soon.

    -Louie

  2. #2
    Senior Member eatmadustch's Avatar
    Join Date
    Nov 2002
    Location
    Switzerland
    Posts
    154
    If you'd tell me how to find out the kernel# I'd gladly help
    EatMaDust


    Stop Microsoft turning into Big Brother!
    http://www.againsttcpa.com

  3. #3
    Junior Member
    Join Date
    Nov 2002
    Location
    Scotland
    Posts
    6
    I have one (Peanut) Linux machine running SOB v1.0.2, and it has seg-faulted occasionally. No obvious reasons why, ie not specifically during submitting - it just decides to have a sulk sometimes

    Never tried any other versions of the client on it, but would be happy to help out testing any new client.

    Ray

  4. #4
    Senior Member wirthi's Avatar
    Join Date
    Apr 2002
    Location
    Pasching.AT.EU
    Posts
    820
    Originally posted by eatmadustch
    If you'd tell me how to find out the kernel# I'd gladly help
    Try:

    cd /proc
    cat version

    Working on SuSe, perhaps on many others

  5. #5
    Senior Member eatmadustch's Avatar
    Join Date
    Nov 2002
    Location
    Switzerland
    Posts
    154
    Louie, you don't want us to test 0.99 as well?
    And do you want us to leave the sb.pns.net, or change it to the IP? Or change it to the proxy sbp.pns.net:80?
    EatMaDust


    Stop Microsoft turning into Big Brother!
    http://www.againsttcpa.com

  6. #6
    Senior Member eatmadustch's Avatar
    Join Date
    Nov 2002
    Location
    Switzerland
    Posts
    154
    SuSE linux 2.4.19-4GB (more /proc/version)
    EDIT: This was a Thunderbird 1GHz, in case you need to know ...
    0.97 -- seems to work (3 tests withough segfaults)
    1.00 -- segfaults on submit (sometimes)
    1.01 -- segfaults on start (if there is no cache and has to download a block [surprisingly only segfaults after caching it to disk]), otherwise on block submission
    1.02 -- works fine (hasn't segfaulted since it was avalable for download, I think )

    rather interesting: 1.0.1 segfaults after caching the packet to disk, 1.0.0 after connecting to server (but only sometimes)

    This was all done withough changing the sclient.conf file, except the username and priority (to low) (no sbp.pns.net with a different port!)

    I would like to test new versions, my e-mail is sobtest.7.eatmadust@xoxy.net (a throw-away-email, so it doesn't matter if spambots find it )

    here's the commented log file

    Code:
    [Sun Jun 15 11:20:09 2003] client process [v0.9.7] invoked
    [Sun Jun 15 11:20:09 2003] priority set to low
    [Sun Jun 15 11:20:09 2003] connecting to server
    [Sun Jun 15 11:20:20 2003] logging into server
    [Sun Jun 15 11:20:21 2003] requesting a block
    [Sun Jun 15 11:20:25 2003] got proth test from server (k=22699, n=233038)
    [Sun Jun 15 11:20:25 2003] server packet cached to disk
    [Sun Jun 15 11:20:57 2003] iteration: 10000/233053 (4.29%) k = 22699  n = 233038
    [Sun Jun 15 11:21:31 2003] iteration: 20000/233053 (8.58%) k = 22699  n = 233038
    [Sun Jun 15 11:22:05 2003] iteration: 30000/233053 (12.87%) k = 22699  n = 233038
    [Sun Jun 15 11:22:39 2003] iteration: 40000/233053 (17.16%) k = 22699  n = 233038
    [Sun Jun 15 11:23:13 2003] iteration: 50000/233053 (21.45%) k = 22699  n = 233038
    [Sun Jun 15 11:23:48 2003] iteration: 60000/233053 (25.75%) k = 22699  n = 233038
    [Sun Jun 15 11:24:22 2003] iteration: 70000/233053 (30.04%) k = 22699  n = 233038
    [Sun Jun 15 11:24:57 2003] iteration: 80000/233053 (34.33%) k = 22699  n = 233038
    [Sun Jun 15 11:25:31 2003] iteration: 90000/233053 (38.62%) k = 22699  n = 233038
    [Sun Jun 15 11:26:06 2003] iteration: 100000/233053 (42.91%) k = 22699  n = 233038
    [Sun Jun 15 11:26:40 2003] iteration: 110000/233053 (47.20%) k = 22699  n = 233038
    [Sun Jun 15 11:27:14 2003] iteration: 120000/233053 (51.49%) k = 22699  n = 233038
    [Sun Jun 15 11:27:48 2003] iteration: 130000/233053 (55.78%) k = 22699  n = 233038
    [Sun Jun 15 11:28:21 2003] iteration: 140000/233053 (60.07%) k = 22699  n = 233038
    [Sun Jun 15 11:28:54 2003] iteration: 150000/233053 (64.36%) k = 22699  n = 233038
    [Sun Jun 15 11:29:31 2003] iteration: 160000/233053 (68.65%) k = 22699  n = 233038
    [Sun Jun 15 11:30:04 2003] iteration: 170000/233053 (72.94%) k = 22699  n = 233038
    [Sun Jun 15 11:30:36 2003] iteration: 180000/233053 (77.24%) k = 22699  n = 233038
    [Sun Jun 15 11:31:08 2003] iteration: 190000/233053 (81.53%) k = 22699  n = 233038
    [Sun Jun 15 11:31:40 2003] iteration: 200000/233053 (85.82%) k = 22699  n = 233038
    [Sun Jun 15 11:32:12 2003] iteration: 210000/233053 (90.11%) k = 22699  n = 233038
    [Sun Jun 15 11:32:45 2003] iteration: 220000/233053 (94.40%) k = 22699  n = 233038
    [Sun Jun 15 11:33:17 2003] iteration: 230000/233053 (98.69%) k = 22699  n = 233038
    [Sun Jun 15 11:33:27 2003] residue: c68c605229df37c
    [Sun Jun 15 11:33:27 2003] completed proth test(k=22699, n=233038): result 3
    [Sun Jun 15 11:33:27 2003] connecting to server
    [Sun Jun 15 11:33:31 2003] logging into server
    [Sun Jun 15 11:33:34 2003] requesting a block
    [Sun Jun 15 11:33:40 2003] got proth test from server (k=33661, n=233064)
    [Sun Jun 15 11:33:40 2003] server packet cached to disk
    [Sun Jun 15 11:34:12 2003] iteration: 10000/233080 (4.29%) k = 33661  n = 233064
    [Sun Jun 15 11:34:44 2003] iteration: 20000/233080 (8.58%) k = 33661  n = 233064
    <snip>
    [Sun Jun 15 11:45:31 2003] iteration: 210000/233080 (90.10%) k = 33661  n = 233064
    [Sun Jun 15 11:46:07 2003] iteration: 220000/233080 (94.39%) k = 33661  n = 233064
    [Sun Jun 15 11:46:42 2003] iteration: 230000/233080 (98.68%) k = 33661  n = 233064
    [Sun Jun 15 11:46:53 2003] residue: dbdf37b65728740
    [Sun Jun 15 11:46:53 2003] completed proth test(k=33661, n=233064): result 3
    [Sun Jun 15 11:46:53 2003] connecting to server
    [Sun Jun 15 11:46:53 2003] logging into server
    [Sun Jun 15 11:46:54 2003] requesting a block
    [Sun Jun 15 11:46:59 2003] got proth test from server (k=5359, n=233086)
    [Sun Jun 15 11:46:59 2003] server packet cached to disk
    [Sun Jun 15 11:47:34 2003] iteration: 10000/233099 (4.29%) k = 5359  n = 233086
    [Sun Jun 15 11:48:10 2003] iteration: 20000/233099 (8.58%) k = 5359  n = 233086
    <snip>
    [Sun Jun 15 11:59:36 2003] iteration: 210000/233099 (90.09%) k = 5359  n = 233086
    [Sun Jun 15 12:00:12 2003] iteration: 220000/233099 (94.38%) k = 5359  n = 233086
    [Sun Jun 15 12:00:48 2003] iteration: 230000/233099 (98.67%) k = 5359  n = 233086
    [Sun Jun 15 12:00:59 2003] residue: 3aef5f33539a36c
    [Sun Jun 15 12:00:59 2003] completed proth test(k=5359, n=233086): result 3
    [Sun Jun 15 12:00:59 2003] connecting to server
    [Sun Jun 15 12:00:59 2003] logging into server
    [Sun Jun 15 12:01:00 2003] requesting a block
    [Sun Jun 15 12:01:05 2003] got proth test from server (k=21181, n=233228)
    [Sun Jun 15 12:01:05 2003] server packet cached to disk
    [Sun Jun 15 12:01:41 2003] iteration: 10000/233243 (4.29%) k = 21181  n = 233228
    [Sun Jun 15 12:02:16 2003] iteration: 20000/233243 (8.57%) k = 21181  n = 233228
    
    ctrl-c
    
    [Sun Jun 15 12:03:19 2003] client process [v1.0.0] invoked
    [Sun Jun 15 12:03:19 2003] priority set to low
    [Sun Jun 15 12:03:19 2003] got k and n from cache
    
    ctrl-c
    
    [Sun Jun 15 12:03:40 2003] client process [v1.0.0] invoked
    [Sun Jun 15 12:03:40 2003] priority set to low
    [Sun Jun 15 12:03:40 2003] connecting to server
    [Sun Jun 15 12:03:40 2003] logging into server
    [Sun Jun 15 12:03:41 2003] requesting a block
    [Sun Jun 15 12:03:49 2003] got proth test from server (k=28433, n=233233)
    [Sun Jun 15 12:03:49 2003] server packet cached to disk
    [Sun Jun 15 12:04:19 2003] iteration: 10000/233248 (4.29%) k = 28433  n = 233233
    [Sun Jun 15 12:04:49 2003] iteration: 20000/233248 (8.57%) k = 28433  n = 233233
    <snip>
    [Sun Jun 15 12:14:55 2003] iteration: 220000/233248 (94.32%) k = 28433  n = 233233
    [Sun Jun 15 12:15:25 2003] iteration: 230000/233248 (98.61%) k = 28433  n = 233233
    [Sun Jun 15 12:15:35 2003] residue: C1524514E641C02A
    [Sun Jun 15 12:15:35 2003] completed proth test(k=28433, n=233233): result 3
    [Sun Jun 15 12:15:35 2003] connecting to server
    [Sun Jun 15 12:15:36 2003] logging into server
    [Sun Jun 15 12:15:37 2003] requesting a block
    [Sun Jun 15 12:15:41 2003] got proth test from server (k=27653, n=233313)
    [Sun Jun 15 12:15:41 2003] server packet cached to disk
    [Sun Jun 15 12:16:12 2003] iteration: 10000/233328 (4.29%) k = 27653  n = 233313
    [Sun Jun 15 12:16:42 2003] iteration: 20000/233328 (8.57%) k = 27653  n = 233313
    [Sun Jun 15 12:17:12 2003] iteration: 30000/233328 (12.86%) k = 27653  n = 233313
    <snip>
    [Sun Jun 15 12:25:47 2003] iteration: 200000/233328 (85.72%) k = 27653  n = 233313
    [Sun Jun 15 12:26:17 2003] iteration: 210000/233328 (90.00%) k = 27653  n = 233313
    [Sun Jun 15 12:26:47 2003] iteration: 220000/233328 (94.29%) k = 27653  n = 233313
    [Sun Jun 15 12:27:17 2003] iteration: 230000/233328 (98.57%) k = 27653  n = 233313
    [Sun Jun 15 12:27:27 2003] residue: 17B1DAAF6CA93C96
    [Sun Jun 15 12:27:27 2003] completed proth test(k=27653, n=233313): result 3
    [Sun Jun 15 12:27:27 2003] connecting to server
    
    ***SEGMENTATION FAULT***
    
    [Sun Jun 15 12:33:08 2003] client process [v1.0.1] invoked
    [Sun Jun 15 12:33:08 2003] priority set to low
    [Sun Jun 15 12:33:08 2003] connecting to server
    [Sun Jun 15 12:33:08 2003] logging into server
    [Sun Jun 15 12:33:09 2003] requesting a block
    [Sun Jun 15 12:33:13 2003] got proth test from server (k=10223, n=233321)
    [Sun Jun 15 12:33:13 2003] server packet cached to disk
    
    ***SEGMENTATION FAULT***
    
    [Sun Jun 15 12:33:29 2003] client process [v1.0.1] invoked
    [Sun Jun 15 12:33:29 2003] priority set to low
    [Sun Jun 15 12:33:29 2003] got k and n from cache
    
    ctrl-c
    
    [Sun Jun 15 12:33:55 2003] client process [v1.0.1] invoked
    [Sun Jun 15 12:33:55 2003] priority set to low
    [Sun Jun 15 12:33:55 2003] connecting to server
    [Sun Jun 15 12:33:56 2003] logging into server
    [Sun Jun 15 12:33:56 2003] requesting a block
    [Sun Jun 15 12:34:01 2003] got proth test from server (k=67607, n=233451)
    [Sun Jun 15 12:34:01 2003] server packet cached to disk
    
    ***SEGMENTATION FAULT***
    
    [Sun Jun 15 12:34:44 2003] client process [v1.0.0] invoked
    [Sun Jun 15 12:34:44 2003] priority set to low
    [Sun Jun 15 12:34:44 2003] connecting to server
    [Sun Jun 15 12:34:45 2003] logging into server
    [Sun Jun 15 12:34:45 2003] requesting a block
    [Sun Jun 15 12:34:50 2003] got proth test from server (k=27653, n=233457)
    [Sun Jun 15 12:34:50 2003] server packet cached to disk
    
    ctrl-c (was just to see if 1.0.0 also segfaults on "cache to disk"/starting without cache)
    
    [Sun Jun 15 12:35:50 2003] client process [v1.0.1] invoked
    [Sun Jun 15 12:35:50 2003] priority set to low
    [Sun Jun 15 12:35:50 2003] got k and n from cache
    [Sun Jun 15 12:36:20 2003] iteration: 10000/233472 (4.28%) k = 27653  n = 233457
    [Sun Jun 15 12:36:50 2003] iteration: 20000/233472 (8.57%) k = 27653  n = 233457
    <snip>
    [Sun Jun 15 12:47:05 2003] iteration: 220000/233472 (94.23%) k = 27653  n = 233457
    [Sun Jun 15 12:47:34 2003] iteration: 230000/233472 (98.51%) k = 27653  n = 233457
    [Sun Jun 15 12:47:45 2003] residue: F34342CE012CF835
    [Sun Jun 15 12:47:45 2003] completed proth test(k=27653, n=233457): result 3
    [Sun Jun 15 12:47:45 2003] connecting to server
    [Sun Jun 15 12:47:46 2003] logging into server
    [Sun Jun 15 12:47:47 2003] requesting a block
    [Sun Jun 15 12:47:52 2003] got proth test from server (k=10223, n=233477)
    [Sun Jun 15 12:47:52 2003] server packet cached to disk
    
    ***SEGMENTATION FAULT***
    Last edited by eatmadustch; 06-15-2003 at 08:04 AM.
    EatMaDust


    Stop Microsoft turning into Big Brother!
    http://www.againsttcpa.com

  7. #7
    Senior Member eatmadustch's Avatar
    Join Date
    Nov 2002
    Location
    Switzerland
    Posts
    154
    On my other linux (on VMWare, a "self made" linux), 2.4.20

    all 4 versions segfault on connecting to server (at startup). I don't know if this has anything to do with VMWare, I actually just installed this linux to see if it works.

    Code:
    [Sun Jun 15 14:04:21 2003] client process [v0.9.7] invoked
    [Sun Jun 15 14:04:21 2003] priority set to low
    [Sun Jun 15 14:04:21 2003] connecting to server
    ***segfault***
    [Sun Jun 15 14:05:39 2003] client process [v1.0.0] invoked
    [Sun Jun 15 14:05:39 2003] priority set to low
    [Sun Jun 15 14:05:39 2003] connecting to server
    ***segfault***
    [Sun Jun 15 14:06:32 2003] client process [v1.0.1] invoked
    [Sun Jun 15 14:06:32 2003] priority set to low
    [Sun Jun 15 14:06:32 2003] connecting to server
    ***segfault***
    [Sun Jun 15 14:06:49 2003] client process [v1.0.2] invoked
    [Sun Jun 15 14:06:49 2003] priority set to low
    [Sun Jun 15 14:06:49 2003] connecting to server
    ***segfault***
    EatMaDust


    Stop Microsoft turning into Big Brother!
    http://www.againsttcpa.com

  8. #8
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    got a ver. 1.02 running smoothly on an 2.4.20-gentoo-r2 (P3-450 MHz) crashed once, but that was fixed changing the IP, no crash since, but a few hangs, probably server related.

    got 6 ver 1.02 running allright on some old Red-Hat version (will find excact version, next time I have access to them(edit Linux version 2.2.12-20)). They crash or hang from time to time, but not very often something aroun.d every 24 hours (P3 ca. 450 MHz) (edit: no crashes or hangs during this weekend)

    I don't think any of the computers have had problems submitting finished tests, but are nor all that sure about the 6 that I don't have frequent acces to.
    Last edited by Frodo42; 06-16-2003 at 04:43 AM.

  9. #9
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    Just browsing my a gigantic log file and come upon this rather anoying thing, whether it was the server or the client I don't know:

    [Wed May 7 07:47:53 2003] n.high = 3420762 . 4 blocks left in test
    [Wed May 7 08:17:20 2003] iteration: 3430000/3483074 (98.48%) k = 55459 n = 3483058
    [Wed May 7 08:49:12 2003] iteration: 3440000/3483074 (98.76%) k = 55459 n = 3483058
    [Wed May 7 08:53:44 2003] logging into server
    [Wed May 7 08:54:05 2003] login unsuccessful -- check your username
    [Wed May 7 09:21:35 2003] iteration: 3450000/3483074 (99.05%) k = 55459 n = 3483058
    [Wed May 7 09:53:28 2003] iteration: 3460000/3483074 (99.34%) k = 55459 n = 3483058
    [Wed May 7 09:59:59 2003] logging into server
    [Wed May 7 10:00:02 2003] login successful
    [Wed May 7 10:00:22 2003] server had no record of proth test, test abandoned
    [Wed May 7 10:00:22 2003] connecting to server
    [Wed May 7 10:00:25 2003] logging into server
    [Wed May 7 10:00:25 2003] requesting a block

  10. #10
    Junior Member bagleyd's Avatar
    Join Date
    Mar 2003
    Location
    Northern New Jersey - US
    Posts
    9
    hi

    0.97 seems to work for me did not heavily test.
    1.00 works but it occassionally gets stuck (this has been reported by others) it is usable though when put in a while loop.
    while (true) do
    ./sb sclient.conf
    done
    Occasionally this may get stuck too... as it waits for something.... giving it a control-C sends it on its merry way again.

    1.01 and 1.02
    [Sun Jun 15 21:17:33 2003] client process [v1.0.2] invoked
    [Sun Jun 15 21:17:33 2003] priority set to idle
    [Sun Jun 15 21:17:33 2003] connecting to server
    [Sun Jun 15 21:17:33 2003] logging into server
    [Sun Jun 15 21:17:33 2003] requesting a block
    [Sun Jun 15 21:17:37 2003] got proth test from server (k=55459, n=3889018)
    [Sun Jun 15 21:17:37 2003] server packet cached to disk
    Segmentation fault

    I would like to be a tester. bagleyd@tux.org
    Also if you can change the name to sob that might cause less confusion as there is already a sb program on linux. Then again, changing the name will cause confusion now too.
    Cheers,
    David Bagley

  11. #11
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    oups forget that
    I'm not one of these frequent crashers but nevertheless I will gladly test the new version on my Geontoo-version

    frodo42(at)get2net.dk (at)=@

  12. #12
    Junior Member
    Join Date
    Jun 2003
    Location
    Near Philadelphia, PA, USA
    Posts
    6
    My computer info:

    os[Linux 2.4.20-4GB-athlon i686] (SuSE 8.2 Pro)
    cpu[AMD Athlon(tm) XP 1800+ @ 1.524 GHz]
    mem[Physical: 35/502MB Free | Swap 1007/1019MB Free]
    up[3h 14m 50s]
    disk[/: 11.10/17.55GB Free | /windows/C: 7.68/18.61GB Free | /dev/shm: 251/251MB Free]
    video[nVidia Corporation NV25 [GeForce4 Ti4200] (rev 163)]
    sound[C-Media Electronics Inc CM8738 (rev 16)]

    No matter what I do, and no matter the version I get this:

    [Mon Jun 16 22:09:36 2003] client process [v1.0.2] invoked
    [Mon Jun 16 22:09:36 2003] priority set to idle
    [Mon Jun 16 22:09:36 2003] connecting to server
    Segmentation fault

  13. #13
    Senior Member eatmadustch's Avatar
    Join Date
    Nov 2002
    Location
    Switzerland
    Posts
    154
    that's weird ... I have SuSE (actually I have 8.1 Pro), my kernel version is just very slightly older, I also have an AMD (Thunderbird-1GHz), but there it works fine (at least 0.9.7 and 1.0.2)! I wonder what it is, that causes the segfaults (cpu?, kernel?, linux distro?).
    EatMaDust


    Stop Microsoft turning into Big Brother!
    http://www.againsttcpa.com

  14. #14
    Hi

    I'am running the SoB client for a few tays. After day 2 the client stopped withnout any error messages, but the CPU load gone bach to 0% (idle=100%).
    After stopping the client with Strg+C and restating it is running witout any problens.

    System: Athlon 650, Suse 7.1
    Client 1.02

    [Mon Jun 16 20:57:45 2003] logging into server
    [Mon Jun 16 20:57:46 2003] login successful
    [Mon Jun 16 20:57:46 2003] n.high = 844611 . 184 blocks left in test
    [Mon Jun 16 21:09:00 2003] iteration: 850000/3885299 (21.88%) k = 5359 n = 3885
    [Mon Jun 16 21:29:51 2003] iteration: 860000/3885299 (22.13%) k = 5359 n = 3885
    [Mon Jun 16 21:32:18 2003] logging into server
    [Mon Jun 16 21:32:19 2003] login successful
    [Mon Jun 16 21:32:19 2003] n.high = 861172 . 183 blocks left in test
    [Mon Jun 16 21:50:47 2003] iteration: 870000/3885299 (22.39%) k = 5359 n = 3885
    (*** hanging, stopped and restarted here ***)
    [Mon Jun 16 23:21:53 2003] client process [v1.0.2] invoked
    [Mon Jun 16 23:21:53 2003] priority set to low
    [Mon Jun 16 23:21:53 2003] got k and n from cache
    [Mon Jun 16 23:21:53 2003] restarting proth test from cache (k=5359, n=3885286)
    [Mon Jun 16 23:31:27 2003] logging into server
    [Mon Jun 16 23:31:28 2003] login successful
    [Mon Jun 16 23:31:28 2003] n.high = 877733 . 182 blocks left in test
    [Mon Jun 16 23:36:15 2003] iteration: 880000/3885299 (22.65%) k = 5359 n = 3885
    [Mon Jun 16 23:57:12 2003] iteration: 890000/3885299 (22.91%) k = 5359 n = 3885
    [Tue Jun 17 00:06:13 2003] logging into server
    [Tue Jun 17 00:06:14 2003] login successful

    Jürgen

  15. #15
    your kernel version you usualy request with
    Code:
    /bin/uname -r
    kernel: 2.4.20
    distro: gentoo
    gcc: 3.2.2
    glibc: 2.3.1


    sb97: runs

    sb100: seg faults:
    Code:
    [Tue Jun 17 20:06:13 2003] client process [v1.0.0] invoked
    [Tue Jun 17 20:06:13 2003] priority set to idle
    [Tue Jun 17 20:06:13 2003] connecting to server
    [Tue Jun 17 20:06:13 2003] logging into server
    [Tue Jun 17 20:06:14 2003] requesting a block
    [Tue Jun 17 20:06:18 2003] got proth test from server (k=5359, n=241990)
    .
    .
    .
    [Tue Jun 17 20:13:53 2003] requesting a block
    Segmentation fault
    happens after the first block was processed and it starts to fetch the next.

    sb101: runs

    sb102: runs


    ill gladly help testing the new client, vinz@N0SPAM@alita.cc

    btw, any clue what the file
    Code:
    487368 Jun 17 19:46 /z3898748
    is? when i run sb as a daemon, it gets created. its full of binary... i have logging to console off.
    also, it creates /cache.
    can you make a "workdir" option for the config file?



    Last edited by alita; 06-17-2003 at 02:34 PM.

  16. #16
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    Alita have you tried chaning the server adress like tihs:
    change the name of the server from ServerAddress sb.pns.net
    to
    ServerAddress 216.163.34.105
    (from this thread

    That worked for me with the exact same Gentoo-version.
    If it does not work for you then it must be something hardware related as I have the same Gentoo version running 1.02 perfectly.

  17. #17
    Frodo, read more exactly; 1.02 runs fine (except that it dies after a few hours, but that should be fixed in 1.10 as Louie said)

    1.00 is seg faulting.

  18. #18
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    oops will try to improve, but I'm a sloppy scientist.

    BTW the Linux production seems to be drobbing (daily production is overtaken by Wirus98 ) so the next version of the Linux client is needed.

    I have a feeling that my 7 Linux machines are not producing all that they could and are stalling after some hours, but since I only see the stats most of the time I can't really be sure cause I'm not present when they work.

  19. #19
    I'd also like to help out in the testing. I'm running 1.0.2 and every few days or so, the client just stalls, no seg faults or anything. Running slack 9.0 w/ 2.4.20

  20. #20
    1.0.2 Will crash on Mandrake 9.1 ( Kernels 2.4.19 to 2.4.21, glibc 2.3.1 ) if a hostname is specified instead of an IP. Immediate segfault on startup.

    1.0.2 Runs fine on Debian Woody (3.0 stable) (2.4.18-bf2.4, glibc 2.2.5) with both hostname and IP.

    1.0.1 and 1.0.0 Crash after a period of time with Mandrake 9.0 (2.4.19 (if i remember right), glibc 2.2.5), Redhat 7.2 (2.4.18-27.7.xsmp, glibc 2.2.4), and possibly Debian Woody (as above -- "possibly" because these machines are slow and take a couple weeks per test). The crashes happened with 0 to 5 or so tests (not blocks) completed.

    The segfault on startup with resolving the hostname appears to be related to glibc 2.3.x, as Redhat 8 also uses this version.

    The two issues that I see in 1.0.2 are the TCP/IP ports in CLOSE_WAIT (which are fixed in 1.1.0?), and that the client will drop to 0% usage after some time. I've changed the RetryWait to 60, and I have yet to see the client drop to 0% since. The RetryWait setting change maybe coincidence.

    I still have access to the Redhat 7.2 machines, but I am not allowed to run long-running background processes any longer. I would be able to run them via cron at night if a convenient method of client shutdown where available (like checking to see if a run file still exists every minute, and shutting down if it is deleted).

    The others are all at home.

  21. #21
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    I still have access to the Redhat 7.2 machines, but I am not allowed to run long-running background processes any longer. I would be able to run them via cron at night if a convenient method of client shutdown where available (like checking to see if a run file still exists every minute, and shutting down if it is deleted).
    Well I use Killall sb to stop the client, it works, but it's not the the best way. A killall command could be cron'ed i guess (i use at, since cron don't run on the computers i use)

  22. #22
    Hater of webboards
    Join Date
    Feb 2003
    Location
    København, Denmark
    Posts
    205
    Originally posted by shifted

    possibly" because these machines are slow and take a couple weeks per test). The crashes happened with 0 to 5 or so tests (not blocks) completed.

    I still have access to the Redhat 7.2 machines, but I am not allowed to run long-running background processes any longer. I would be able to run them via cron at night if a convenient method of client shutdown where available (like checking to see if a run file still exists every minute, and shutting down if it is deleted).
    Couldn't you just make them do supersecret tests?

    .Henrik

  23. #23
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    Just got a new P4 2.4 GHz that i also installed with the newest version of Gentoo-Linux, and surprisingly it segfaults SoB almost before it starts, my P3-450 with the same Gentoo-Linux version runs SoB smoothly. The only differences are:
    1. The physical system
    2. On my P4 i compiled everything with the P4 flag instead of the P3 flag
    3. Some diffrent packets installed.

    Anyways the new Linux client is still badly needed!

  24. #24
    Dual AMD Opteron 240
    1536 MB RAM
    RioWorks HDAMB motherboard
    Slackware Linux 9.0
    Kernel 2.4.21-ac4 (32-bit)

    Clients 1.0.1 and 1.0.2 segfault immediately when processing starts.

    [Mon Jul 7 17:07:13 2003] client process [v1.0.2] invoked
    [Mon Jul 7 17:07:13 2003] priority set to idle
    [Mon Jul 7 17:07:13 2003] connecting to server
    [Mon Jul 7 17:07:14 2003] logging into server
    [Mon Jul 7 17:07:14 2003] requesting a block
    [Mon Jul 7 17:07:18 2003] got proth test from server (k=55459, n=3980434)
    [Mon Jul 7 17:07:18 2003] server packet cached to disk
    Segmentation fault

    [Mon Jul 7 17:20:34 2003] client process [v1.0.2] invoked
    [Mon Jul 7 17:20:34 2003] priority set to idle
    [Mon Jul 7 17:20:34 2003] got k and n from cache
    Segmentation fault

    Client 1.0.0 has been running since last night without any problems. It did segfault once, but that may have been a fluke. 0.9.7 works fine as well.

    To summarize:

    0.9.7: works
    1.0.0: works
    1.0.1: segfault on startup
    1.0.2: segfault on startup

    I can be contacted at (my user name)@garandnet.net if I can be of any assistance.
    Last edited by Electrode; 07-12-2003 at 07:32 PM.

  25. #25
    Junior Member
    Join Date
    Dec 2002
    Location
    Las Vegas, NV -- N 36° 10' W 115° 17'
    Posts
    3

    linux report

    Debian 3.0 (stable/woody)box (kernal 2.4.21)
    Intel(R) Pentium(R) 4 CPU 2.53GHz, upgraded libc,upgraded libnss.

    v0.97 -- not tested
    v1.00 -- runs, no crashes
    v1.01 -- crashes on startup
    v1.02 -- crashes on startup

    If you want me to try new versions, send me a note at:

    daylight@yipemail.com
    if brute force won't solve your problem, you aren't using enough

  26. #26

    numeric ip fixed it for me

    I was getting a segfault when it would try to fetch blocks running 1.02.

    Mandrake 9.1
    2.4.21
    gcc 3.2.2

    When I went into the config file and changed the block server into the numeric ip addredd instead of sb.... the problem seems to have gone away. It has run for 24+ hours now whereas it used to crash at the end of the first block that it completed.

    --chris

  27. #27
    i wanted to thank everyone for their feedback and give you an update.

    i have started the work of fixing up the linux client. it's on the top of my programming list. however, I am going to California for the rest of August and will not be able to work on it.

    in mid/late-september i'd expect to have something ready for people to test. email me if you want to test. i'll try to email it to those of you who posted your emails on this thread but i'm certainly not going to chase people down though PM's. so far a total of *ZERO* people have directly emailed me to volunteer. i was actually surprised. even this thread has surprised me by how many people have no problems. those are the main reasons i didn't work on it sooner... the problem with the current client appears even smaller than i imagined and no one contacted me directly. the four of you who did post your emails will definately be getting the new client when i'm ready with it.

    so just to make it clearer, if you have a machine where you can crash the current client and you have the comparison data I asked for in the first post, please email it to lhelm@umich.edu with a subject of "SB LINUX TESTER". I won't respond right away since I'll be gone for at least a month but I'll use the data and keep you in the loop when devel continues. thanks again for your help.

    -Louie

  28. #28
    just to update my situation, about 4-5 weeks ago, the client just stopped having any issues whatsoever. It just goes forever until I kill it. I didn't make any real changes until yesterday, a new motherboard/cpu combo, and even still its goin great. Maybe it had something to do with the work being sent to the clients at the time?

  29. #29
    Junior Member
    Join Date
    Mar 2003
    Location
    St. Louis, Missouri, USA
    Posts
    6

    Seg Faults

    Segmentation Faults -

    Most of my CPU farm is Gentoo Linux based. Recently I've upgrades to the lastest pacakges and now am getting immediate segmentation faults from the 1.02 Linux SB client. Oddly, on one dual PII processor system, the 1st invocation will run fine, but the second (in a seperate directory) dies immediately. Thought it might have something to do with the dual CPU, but noticed sb wouldn't start on my newly rebuilt farm machines as well - same symptom.

    Setiathome seems to be running fine on all of these machines.

    Any thoughts? I was just starting to get back into the stats race when this came up...

    Kevin C.

  30. #30
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    Perhaps some kind of problem with implementing P4 in Gentoo?1.02 crashes segfaults right away on a P4 with newest Gentoo and runs without troubes on a P3, i haven't tried compiling the kernel without the P4-tag on my P4 yet, but I guess I ought to test it.

  31. #31
    Junior Member
    Join Date
    Mar 2003
    Location
    St. Louis, Missouri, USA
    Posts
    6

    Segmentation fault

    Alas, these are P3 and Athlon CPUs. Appreciate the input though.

    Kevin C.


    Originally posted by Frodo42
    Perhaps some kind of problem with implementing P4 in Gentoo?1.02 crashes segfaults right away on a P4 with newest Gentoo and runs without troubes on a P3, i haven't tried compiling the kernel without the P4-tag on my P4 yet, but I guess I ought to test it.

  32. #32
    Junior Member
    Join Date
    Mar 2003
    Location
    St. Louis, Missouri, USA
    Posts
    6

    Segmentation Fault

    Good news!

    Found the hint to use the IP address of sb.pns.net (216.163.34.105) in /etc/sclient.conf instead of the name. For whatever reason, that appears to get things running again.

    Kevin C.

  33. #33
    Senior Member Frodo42's Avatar
    Join Date
    Nov 2002
    Location
    Jutland, Denmark
    Posts
    299
    Is there a way to stop the Linux client in a way that it saves the work it has done?
    I kill the client once every 24 hours (using at, to run the client on these old Linux machines when no one uses them) and as I understand the client saves it's work every 10 min. therefore I loose 5 min. in average every day, times 7 machines that's 35 min. of processorwork every day

    So if this nice exit does exist then tell me how to do it, if not then I would very much like it to be implemented in the next Linux-client made (for example if it recieves a quit or term command, then it saves it's work and exits)
    Last edited by Frodo42; 09-01-2003 at 08:33 AM.

  34. #34
    Junior Member
    Join Date
    Sep 2003
    Location
    Salt Lake City
    Posts
    1

    Mandrake crashes

    I have just tried 97, 99, 100, 101, and 102 on

    Mandrake 9.2 Linux,
    Pentium III
    512MB ram
    Kernel: 2.4.21-0.13mdk

    and all of them segfault right after connecting to server:
    [Mon Sep 1 12:22:44 2003] client process [v0.9.9] invoked
    [Mon Sep 1 12:22:44 2003] priority set to idle
    [Mon Sep 1 12:22:44 2003] connecting to server
    Segmentation fault

    Is it possible to get a copy of the source and I could see if I could tell more precisely where the fault is occurring. Curiously enough I tried the Linux 102 on NetBSD and it ran fine there.

  35. #35
    I had the same problems with all SB clients from 0.97 - 1.02 on my Gentoo box (2.4.20-gentoo-r5).
    They all crashed like this:

    [Tue Sep 2 16:11:36 2003] client process [v1.0.2] invoked
    [Tue Sep 2 16:11:36 2003] priority set to idle
    [Tue Sep 2 16:11:36 2003] connecting to server
    Segmentation fault

    Changing the server adress in the config file to ServerAddress 216.163.34.105 fixed that problem.

    On all my other Suse 8.1 Boxes i had problems with the 1.02 client hanging after network problems. The client would stay in sleep mode forever.

  36. #36
    Junior Member
    Join Date
    Mar 2003
    Location
    St. Louis, Missouri, USA
    Posts
    6
    Its been a couple of weeks now running with the IP address trick. That has helped, but not solved the problem. I'm still getting the occassional segmentation fault, and the occassional sleep forever.

    I've started leaving "top" running with a 300 second refresh and checking the sessions twice a day. With 10 sessions, its not unusual to have to restart 1 or 2 per day.

    Kevin C.

  37. #37
    For those occasional crashes or sleep problems i have written myself a small script which takes care of those problems.

    sbrun.sh

    If anyone in interested in trying it, just go ahead. A little help is included. If you need any help or run into some problems just contact me.

  38. #38
    Member
    Join Date
    Nov 2002
    Location
    Haverhill, MA
    Posts
    76
    i am using red hat 9.0.93 and none of the version of sob works

  39. #39
    Junior Member
    Join Date
    Mar 2003
    Location
    St. Louis, Missouri, USA
    Posts
    6
    Originally posted by Ptah
    For those occasional crashes or sleep problems i have written myself a small script which takes care of those problems.

    sbrun.sh

    If anyone in interested in trying it, just go ahead. A little help is included. If you need any help or run into some problems just contact me.
    This has helped a lot.

    Thank You!

    Kevin C.

  40. #40
    FWIW, I haven't had any crashing/hang problems since using "sbwrap".

    The link was posted at the Ars forums a while ago:
    http://www.its.caltech.edu/~sgeier/code/sbwrap.html

Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •