• OS/2 application stack space

    From Sean Dennis@1:11/200 to All on Thu Nov 28 12:23:10 2002
    Hello, All!

    I'm thinking OS/2 is giving me signs that I may be pushing it too hard... I have a 850Mhz Athlon Thunderbird-powered computer for my BBS with 96MB RAM and two hard drives (OS/2 resides on a 2GB HD and the BBS lives on a 20GB HD). I'm
    running Maximus/2 v3.01 with the Y2K fixes.

    I also use this machine as my main Internet server. I normally have 18-22 programs running at all times for various services, etc. What I'm noticing is some strange problems when callers come in with Max/2. Max will suddenly just drop carrier on them as well as crash when looking at the Fidonet message area list (I carry the entire Fido backbone, about 430+ echoes).

    What I'm thinking is that OS/2 may be running out of room and simply acting strangely. I do have the OS/2 memory kludge in my BIOS turned OFF (I believe so-I can't reboot the machine right now to find out).

    I have 192MB RAM in my other machine that I use heavily. I may switch the RAM from that one over to this computer to see if that would help.

    If anyone has any ideas or suggestions, I'd really appreciate the help.

    Thanks in advance,
    Sean

    // hausmaus@midnightshour.org | http://midnightshour.org | ICQ: 19965647

    --- FleetStreet 1.27.1
    * Origin: Midnight's Hour BBS - Carbondale, IL - midnightshour.org (1:11/200)
  • From Neil Walker@2:250/501 to Sean Dennis on Thu Nov 28 22:50:21 2002
    Hello Sean!

    Thursday November 28 2002 12:23, you wrote to All:

    I also use this machine as my main Internet server.

    What do you mean by Internet server?

    I ran BinkD, an FTP server and a POP3/SMTP server in addition to 3 dial-up nodes and a Telnet node on a P133 with 96MB RAM without problems. I didn't run a Web server, though.

    I normally have 18-22 programs running at all times for various
    services, etc.

    Hmm. Do they really all need to run at once?

    I carry the entire Fido backbone, about 430+ echoes).

    Me too.

    What I'm thinking is that OS/2 may be running out of room and simply acting strangely. I do have the OS/2 memory kludge in my BIOS turned
    OFF (I believe so-I can't reboot the machine right now to find out).

    What size is your swap file at before the crash? How much free disk space is left on the swap file disk?


    Be lucky,

    Neil

    This OS/2 system uptime is 71 days 22:53 hours and 11 seconds
    ... TAGLINE A $300.00 Picture tube will protect a 10 cent fuse by blowing first.

    --- GoldED/2 3.0.1
    * Origin: The Electric Pigeon, Telford, UK (2:250/501)
  • From Sean Dennis@1:11/200 to Neil Walker on Thu Nov 28 19:27:58 2002
    Hello, Neil!

    Replying to a message of Neil Walker to Sean Dennis:

    What do you mean by Internet server?

    Runs all of my Internet-related servers.

    Hmm. Do they really all need to run at once?

    Yes. Here's what I'm running:

    Internet Rex
    GoServe (Gopher server)
    Web/2 (httpd)
    BinkleyTerm/2
    Two Max/2 telnet nodes
    VMODEM
    TIMEXEC (my cron)
    Webmail/2 WebManager (I offer web access to my pop3 BBS accounts)
    Maximus Session Monitor
    Netfinity Network Interface
    Network Messaging
    OpenChat/2 w/three windows
    Desktop
    Warpcenter

    That's what's normally running at all times on my machine.

    What size is your swap file at before the crash? How much free disk
    space is left on the swap file disk?

    Next time the system crashes, I'll try to catch that.

    Later,
    Sean

    // hausmaus@midnightshour.org | http://midnightshour.org | ICQ: 19965647

    --- FleetStreet 1.27.1
    * Origin: Midnight's Hour BBS - Carbondale, IL - midnightshour.org (1:11/200)
  • From Peter Knapper@3:772/1.10 to Sean Dennis on Fri Nov 29 18:51:50 2002
    Hi Sean,

    I'm thinking OS/2 is giving me signs that I may be
    pushing it too hard... I have a 850Mhz Athlon
    Thunderbird-powered computer for my BBS with 96MB RAM
    and two hard drives (OS/2 resides on a 2GB HD and the
    BBS lives on a 20GB HD). I'm running Maximus/2 v3.01
    with the Y2K fixes.

    As a comparison, I have a Celeron 700 with 64Mb, 10GB HD, running Warp 4 FP9 with Binkley/2 2.60, Maximus & Squish 3.01 + Y2k, 2 x PSTN lines, BinkD, DNS, NTP/SNTP server, and this machine runs for a month before I do a regular shutdown/restart. It rarely drops below 24MB memory free, so its not short of resources. Its not used for anything else other than the BBS and has been a pretty stable box since it was built (about 18 months ago). Currently its running 15 Tasks, 39 Processes and 132 threads, but things are quiet as I type.

    Max will suddenly just drop carrier on them

    If Max drops carrier, then the Max log tells you WHY it did this. If you are not seeing any Max messages in the log, then something else is causing the loss
    of carrier.

    as well as crash when looking at the Fidonet message area list (I
    carry the entire Fido backbone, about 430+ echoes).

    If Max crasshes on attempting to access a Message area, then look for soe sort of message area corruption, or a HD Partition corruption, or an HD issue of some sort. Run SQFIXP over all areas and see what that reports.

    Finally, (assuming HPFS on all partitions) after doing EVERYTHING else possible
    first because it can take a heck of a LONG time to complete (like over an hour!!!) on LARGE partitions, run a CHKDSK /F:3 on your message base drive, then all other drives. This mihght dig up a lot of old files, but carefuly check what it finds. This can be a useful in sight into an HD problem that things are getting bad somewhere.

    What I'm thinking is that OS/2 may be running out of
    room and simply acting strangely.

    Is your SWAP file increasing in size, or do you have swapping turned OFF????

    I do have the OS/2
    memory kludge in my BIOS turned OFF (I believe so-I
    can't reboot the machine right now to find out).

    I have never needed that, its always seen the full memory on my machines without using it.

    I have 192MB RAM in my other machine that I use
    heavily. I may switch the RAM from that one over to
    this computer to see if that would help.

    It is remotely possible you __MIGHT__ have a memory problem that OS/2 is runing
    into with rarely used memory. I cursed the day they took parity memory out of the PC world, the BIOS check is NOT a 100% guarantee. Try an external memory checking program and see if it finds anything.

    Side track here... I ran into my first real Memory problem in a LONG time about
    4 weeks ago when I upgraded an OS/2 machine from 128MB to 512MB. The upgrade was fine, but after about 12 hours of error free runing it trapped. The restart
    POST was fine, however CHKDSK found all sorts of corrupt files and just trapped
    again. Eventually, when I backed out to my orignal 128MB, everything was fine, no corrupt HD or anything!!!. It turns out the Motherboard chipset couldn't run
    a full 512MB (2 x 256K sticks) reliably (apparently a known issue with this particular Chipset), however the machine has been running fine on 384Kb ever since I discovered the problem.

    If anyone has any ideas or suggestions, I'd really appreciate the help.

    Thats about all so far, see how that goes.

    Cheers............pk.


    --- Maximus/2 3.01
    * Origin: Another Good Point About OS/2 (3:772/1.10)
  • From Vince Coen@2:257/609 to Sean Dennis on Fri Nov 29 13:57:58 2002
    Hello Sean!

    28 Nov 02 12:23, you wrote to All:

    I'm thinking OS/2 is giving me signs that I may be pushing it too
    hard... I have a 850Mhz Athlon Thunderbird-powered computer for my BBS with 96MB RAM and two hard drives (OS/2 resides on a 2GB HD and the
    BBS lives on a 20GB HD). I'm running Maximus/2 v3.01 with the Y2K
    fixes.

    I use 192mb Ram and I sometimes run out with that using two Max/Bink nodes. I should point out that OS/2 starts getting problematic when free Ram drops below around 90mb.

    Hmm, another point be very carefull when adding Ram. OS/2 does not like mismatched Ram simms i.e., use identical if possible.

    Vince

    --- Linux/Mbse/GoldED+/LNX 1.1.5-20709
    * Origin: Air Applewood, Linux Gateway for Essex (2:257/609)
  • From Gord Hannah@1:17/23.1 to Sean Dennis on Fri Nov 29 09:03:23 2002
    Replying to a message from Sean Dennis 1:11/200 to All,

    About OS/2 application stack space, On Thu Nov 28 2002

    I'm thinking OS/2 is giving me signs that I may be pushing it too

    If anyone has any ideas or suggestions, I'd really appreciate the
    help.

    Do you have go if not get it from Hobbes or let me know I will send it to you..

    Go:

    GO! v1.5 - (c) 1993-95 by Carsten Wimmer <cawim@train.oche.de>

    Usage: go <command> [argument]

    Available commands:

    -pl Process List <default> -k Kill a Process
    -lpl Process List with Paths -ka Kill all Instances
    -tl Thread List -cp Check for a Process
    -sl Semaphore List -j Jump to a Process
    -sm Shared Memory -at Application Type
    -ml Module List -ut Machine Uptime
    -mt Module Dependency Tree -df Free Space on Drives
    -lh Long, detailed Help
    -la GO! License Agreement

    go > golist.txt
    GO! v1.5 - (c) 1993-95 by Carsten Wimmer <cawim@train.oche.de>

    List of Processes:

    P-ID PPID Session Thr Prio CPU Time Name
    ---- ---- ------- --- ---- -------------- ------------------------------------
    24 0 000 Det 1 0300 0:00:00.15 VDOSCTL.EXE
    18 0 000 Det 8 0304 25:17:03.34 CNTRL.EXE
    17 0 000 Det 1 0200 0:00:00.46 LANMSGEX.EXE
    11 0 000 Det 1 031F 0:00:00.06 MIDIDMON.EXE
    10 0 000 Det 1 0200 0:00:00.34 SMSTART.EXE
    15 10 000 Det 1 0200 0:00:00.65 FFST.EXE
    14 10 000 Det 1 021F 0:00:00.90 ERLOGGER.EXE
    13 10 000 Det 2 0200 0:00:00.34 DMISL.EXE
    25 1 001 Sys 22 0200 0:40:38.71 PMSHELL.EXE
    604 25 01B VIO 1 0200 0:00:00.28 CMD.EXE
    609 604 01B VIO 1 0200 0:00:00.00 GO.EXE
    584 25 017 VIO 1 0200 0:00:00.71 CMD.EXE
    587 584 017 VIO 2 0200 0:01:09.62 TIMEDP.EXE
    602 587 017 VIO 1 0200 0:00:00.03 CMD.EXE
    603 602 017 VIO 1 0200 0:00:00.84 Q.EXE
    47 25 018 VIO 1 0200 34:17:17.87 FC.EXE
    46 25 016 PM 1 0200 0:00:08.59 MINICALC.EXE
    45 25 015 PM 2 0200 1:50:55.15 ANACLOCK.EXE
    44 25 014 VIO 1 0200 0:00:00.09 CMD.EXE
    43 25 01C PM 5 040A 3:03:49.15 CCHMAIN.EXE
    38 25 01A VIO 2 0200 0:26:45.93 CRON2.EXE
    37 25 019 PM 1 0200 13:08:23.21 BLANKER.EXE
    31 25 013 PM 2 0200 22:06:19.12 MXCS3.EXE
    30 25 012 PM 2 0200 0:00:00.43 ARTCHRON.EXE
    29 25 011 PM 20 0200 6:05:44.68 PMSHELL.EXE
    28 25 010 PM 5 0200 0:00:14.03 PMSPOOL.EXE
    26 25 000 Sys 3 0300 0:00:00.34 HARDERR.EXE
    2 1 000 VDM 1 031F 0:00:00.03 VDM There are 28 Processes with 91 Threads. This machine's uptime is 4d 13h 54m 18s 781ms.

    Also what is the line in your config.sys dealing with your swapper.dat file?

    Mine is: MEMMAN=SWAP,PROTECT
    SWAPPATH=D:\OS2\SYSTEM 2048 2048

    As I write this today is a very light day.

    System a K62/300 with 320MB mem

    Hope this helps. Keep us posted.

    We are a fine board trying to make it better.
    http://www.pris.bc.ca/ghannah
    ghannah@pris.bc.ca
    Cheers! Gord
    -=Team OS/2=-
    --- timEd/2 1.10.y2k+
    * Origin: Marsh BBS (c), Dawson Creek, BC Canada (1:17/23.1)
  • From Sean Dennis@1:11/200 to Vince Coen on Fri Nov 29 22:48:50 2002
    Hello, Vince!

    Replying to a message of Vince Coen to Sean Dennis:


    I use 192mb Ram and I sometimes run out with that using two Max/Bink nodes. I should point out that OS/2 starts getting problematic when
    free Ram drops below around 90mb.

    I was thinking that... that means I must be stretching the living crap out of it then!

    Hmm, another point be very carefull when adding Ram. OS/2 does not
    like mismatched Ram simms i.e., use identical if possible.

    That I know from experience. I think that's part of the problem too. I have no clue where this RAM in the BBS machine came from.


    Later,
    Sean

    // hausmaus@midnightshour.org | http://midnightshour.org | ICQ: 19965647

    --- FleetStreet 1.27.1
    * Origin: Midnight's Hour BBS - Carbondale, IL - midnightshour.org (1:11/200)
  • From Dale A Cook@1:109/347 to Sean Dennis on Sun Dec 15 23:34:08 2002
    Hello, Neil!

    Replying to a message of Neil Walker to Sean Dennis:

    What do you mean by Internet server?

    Runs all of my Internet-related servers.

    Hmm. Do they really all need to run at once?

    Yes. Here's what I'm running:

    Internet Rex
    GoServe (Gopher server)
    Web/2 (httpd)
    BinkleyTerm/2
    Two Max/2 telnet nodes
    VMODEM
    TIMEXEC (my cron)
    Webmail/2 WebManager (I offer web access to my pop3 BBS accounts)
    Maximus Session Monitor
    Netfinity Network Interface
    Network Messaging
    OpenChat/2 w/three windows
    Desktop
    Warpcenter

    That's what's normally running at all times on my machine.

    What size is your swap file at before the
    crash? How much free disk
    space is left on the swap file disk?

    Next time the system crashes, I'll try to catch that.

    Later,
    Sean

    // hausmaus@midnightshour.org |
    http://midnightshour.org | ICQ: 19965647

    Hey Sean: I can donate some xtra ram if you like? Ram aka Memory is
    like CASH::: You can't ever have TOO MUCH of it !!

    --- Maximus/2 2.02
    * Origin: OS/2 Shareware BBS, telnet://bbs.os2bbs.com (1:109/347)
  • From Sean Dennis@1:11/200 to Dale A Cook on Mon Dec 16 23:38:34 2002
    Hello, Dale!

    Replying to a message of Dale A Cook to Sean Dennis:

    Hey Sean: I can donate some xtra ram if you like? Ram aka Memory is like CASH::: You can't ever have TOO MUCH of it !!

    I don't have any money now for any, but I'd be happy to get whatever I could. Thanks for the generous offer.

    Later,
    Sean

    // hausmaus@midnightshour.org | http://midnightshour.org | ICQ: 19965647

    --- FleetStreet 1.27.1
    * Origin: Midnight's Hour BBS - Carbondale, IL - midnightshour.org (1:11/200)
  • From Mike Luther@1:117/3001 to Sean Dennis on Tue Dec 17 07:41:56 2002
    But, so said .. y'all

    Hello, Dale!

    Replying to a message of Dale A Cook to Sean Dennis:

    Hey Sean: I can donate some xtra ram
    if you like? Ram aka
    Memory is like CASH::: You can't ever have TOO MUCH of it !!

    I don't have any money now for any, but I'd be happy
    to get whatever I could. Thanks for the generous
    offer.

    Be careful about mixing memory sticks from different sources in OS/2. It don't
    like timing and tiny differences between different chunks and can present some very hard to diagnose glitches if you do this. I think from what I've read, these issues depend on how fast the CPU company is that you keep as well. Fast
    company can make things worse.. ;)

    Sorta like just because you have a neat color printer these days and it's easy to print money to buy a stick or two..

    doesn't mean that the devil is not in the details..

    chuckle.


    Sleep well; OS/2's still awake! ;)

    Mike @ 1:117/3001




    --- Maximus/2 3.01
    * Origin: Ziplog Public Port (1:117/3001)
  • From Peter Knapper@3:772/1 to Dale A Cook on Wed Dec 25 13:42:56 2002
    Hi Dale,

    Hey Sean: I can donate some xtra ram if you
    like? Ram aka Memory is like CASH::: You can't
    ever have TOO MUCH of it !!

    Oh yes you can.......;-) I had a motherboard that would work fine with anything
    up to 384MB RAM, but became intermittently unstable at 512MB. Apparently it was
    a problem with one particular memory management chipset the board used.....;-)

    Cheers................pk.


    --- Maximus/2 3.01
    * Origin: === NZCC Maxie BBS. Ak, NZ +64 9 444-0989 === (3:772/1)