• Multiserver

    From Alterego@VERT/ALTERANT to All on Wednesday, September 25, 2019 09:45:51
    So I thought I'd have a go at spinning up another server of SBBS - that would represent (for now) node 4.

    I cloned sbbs.ini and called it sbbs.hostname.ini (where hostname is the hostname of the new server). I've pretty much disabled some things, and have firstnode = 4, lastnode =4

    I cloned services.ini and called it services.hostname.ini - I've basically done enabled=false to many services.

    Started SBBS - and it all looks good.

    However, when I login, it cannot find any of the screens - the logs shows:

    9/25 09:31:07 term Node 4 !Error 2 (No such file or directory) opening: menu/../answer.asc (so no welcome screen)
    9/25 09:39:15 term Node 4 <Guest> !Error 2 (No such file or directory) opening: menu/main.asc (so no menu).

    The second server can see the exact same paths as my primary server (which renders the welcome and menus fine).

    My SBBS is in /opt/sbbs - and main.asc is in /opt/sbbs/text/menu/main.asc

    My answer is answer.msg and its in /opt/sbbs/text/answer.msg

    What have I messed up?
    ...ëîå*

    ... A dog is the only thing on earth that loves you more than you love yourself.

    ---
    þ Synchronet þ Alterant | an SBBS in Docker on Pi!
  • From Digital Man@VERT to Alterego on Tuesday, September 24, 2019 17:23:25
    Re: Multiserver
    By: Alterego to All on Wed Sep 25 2019 09:45 am

    So I thought I'd have a go at spinning up another server of SBBS - that would represent (for now) node 4.

    I cloned sbbs.ini and called it sbbs.hostname.ini (where hostname is the hostname of the new server). I've pretty much disabled some things, and have firstnode = 4, lastnode =4

    I cloned services.ini and called it services.hostname.ini - I've basically done enabled=false to many services.

    Started SBBS - and it all looks good.

    However, when I login, it cannot find any of the screens - the logs shows:

    9/25 09:31:07 term Node 4 !Error 2 (No such file or directory) opening: menu/../answer.asc (so no welcome screen)
    9/25 09:39:15 term Node 4 <Guest> !Error 2 (No such file or directory) opening: menu/main.asc (so no menu).

    The second server can see the exact same paths as my primary server (which renders the welcome and menus fine).

    My SBBS is in /opt/sbbs - and main.asc is in /opt/sbbs/text/menu/main.asc

    My answer is answer.msg and its in /opt/sbbs/text/answer.msg

    What have I messed up?

    It sounds like your node 4 "Text Directory" is not set to "../text/" (the default).

    digital man

    Synchronet "Real Fact" #13:
    Synchronet was the first BBS software to ship with internal QWK networking. Norco, CA WX: 89.0øF, 38.0% humidity, 13 mph E wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • From Alterego@VERT/ALTERANT to Digital Man on Wednesday, September 25, 2019 11:15:27
    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 05:23 pm

    9/25 09:31:07 term Node 4 !Error 2 (No such file or directory) opening: menu/../answer.asc (so no welcome screen)
    9/25 09:39:15 term Node 4 <Guest> !Error 2 (No such file or directory) opening: menu/main.asc (so no menu).
    It sounds like your node 4 "Text Directory" is not set to "../text/" (the default).

    Relative to SBBSCTRL environment variable?

    If yes, then its there (same as main BBS), ie:

    root@ansitex:/opt/sbbs# env|grep SBBS
    SBBSCTRL=/opt/sbbs/ctrl
    SBBSEXEC=/opt/sbbs/exec
    root@ansitex:/opt/sbbs# ls -al $SBBSCTRL/../text/answer.msg
    -rw------- 1 500 500 3275 Aug 19 09:54 /opt/sbbs/ctrl/../text/answer.msg root@ansitex:/opt/sbbs# ls -al $SBBSCTRL/../text/menu/main.asc
    -rw-r--r-- 1 root root 1326 Sep 25 11:12 /opt/sbbs/ctrl/../text/menu/main.asc

    And this is running as 'root' - so its not a permission issue. (but you can see main.asc is 644 anyway).
    ...ëîå*

    ... You don't get once-in-a-lifetime offers like this every day.

    ---
    þ Synchronet þ Alterant | an SBBS in Docker on Pi!
  • From Digital Man@VERT to Alterego on Tuesday, September 24, 2019 18:34:39
    Re: Multiserver
    By: Alterego to Digital Man on Wed Sep 25 2019 11:15 am

    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 05:23 pm

    9/25 09:31:07 term Node 4 !Error 2 (No such file or directory) opening: menu/../answer.asc (so no welcome screen)
    9/25 09:39:15 term Node 4 <Guest> !Error 2 (No such file or directory) opening: menu/main.asc (so no menu).
    It sounds like your node 4 "Text Directory" is not set to "../text/" (the default).

    Relative to SBBSCTRL environment variable?

    If yes, then its there (same as main BBS), ie:

    root@ansitex:/opt/sbbs# env|grep SBBS
    SBBSCTRL=/opt/sbbs/ctrl
    SBBSEXEC=/opt/sbbs/exec
    root@ansitex:/opt/sbbs# ls -al $SBBSCTRL/../text/answer.msg
    -rw------- 1 500 500 3275 Aug 19 09:54 /opt/sbbs/ctrl/../text/answer.msg root@ansitex:/opt/sbbs# ls -al $SBBSCTRL/../text/menu/main.asc
    -rw-r--r-- 1 root root 1326 Sep 25 11:12 /opt/sbbs/ctrl/../text/menu/main.asc

    And this is running as 'root' - so its not a permission issue. (but you can see main.asc is 644 anyway).

    I mean, in SCFG->Nodes->Node 4->Advanced, what do you have the "Text Directory" set to?

    digital man

    Synchronet "Real Fact" #81:
    Vertrauen has had the FidoNet node number 1:103/705 since 1992.
    Norco, CA WX: 84.1øF, 41.0% humidity, 5 mph ESE wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • From Alterego@VERT/ALTERANT to Digital Man on Wednesday, September 25, 2019 12:08:55
    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 06:34 pm

    I mean, in SCFG->Nodes->Node 4->Advanced, what do you have the "Text Directory" set to?

    Ahh (I knew I'd seen that somewhere).

    Node 4's details were completely blank? (I had not touched them before).

    But what I had tryed, when trying to insert (my earlier message about no INS key), I had discovered CTRL-C and CTRL-V (thinking it might append a new entry) - it infact cause scfg to core dump.

    When I tried it a second time (and again just now) copying "Node 3" to "Node 4", it appeared to not do anything.

    I've set the text dir now - and it works - thank you.
    ...ëîå*

    ... A raisin dropped in a glass of fresh champagne will bounce up and down continuously from the bottom of the glass to the top.

    ---
    þ Synchronet þ Alterant | an SBBS in Docker on Pi!
  • From Digital Man@VERT to Alterego on Tuesday, September 24, 2019 21:01:19
    Re: Multiserver
    By: Alterego to Digital Man on Wed Sep 25 2019 12:08 pm

    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 06:34 pm

    I mean, in SCFG->Nodes->Node 4->Advanced, what do you have the "Text Directory" set to?

    Ahh (I knew I'd seen that somewhere).

    Node 4's details were completely blank? (I had not touched them before).

    That's weird and unexpected.

    But what I had tryed, when trying to insert (my earlier message about no INS key), I had discovered CTRL-C and CTRL-V (thinking it might append a new entry) - it infact cause scfg to core dump.

    That's probably how the node's settings got all zeroed out.

    Can you get a backtrace from that core dump? That'd be very helpful in fixing that bug:
    http://wiki.synchro.net/howto:gdb

    When I tried it a second time (and again just now) copying "Node 3" to "Node 4", it appeared to not do anything.

    If the settings between node 3 and 4 were the same, then it could appear that way. But if you make a change (say, the phone number) and then copy paste from one node to another, the settings should change.

    I've set the text dir now - and it works - thank you.

    Okay, cool. Now if I could fix that SCFG bug... :-)

    digital man

    This Is Spinal Tap quote #11:
    Nigel Tufnel: No. no. That's it, you've seen enough of that one.
    Norco, CA WX: 79.5øF, 37.0% humidity, 0 mph SSW wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • From Alterego@VERT/ALTERANT to Digital Man on Wednesday, September 25, 2019 15:20:13
    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 09:01 pm

    But what I had tryed, when trying to insert (my earlier message about no INS key), I had discovered CTRL-C and CTRL-V (thinking it might append a new entry) - it infact cause scfg to core dump.
    That's probably how the node's settings got all zeroed out.
    Can you get a backtrace from that core dump? That'd be very helpful in fixing that bug:

    Reading symbols from ../exec/scfg...(no debugging symbols found)...done.
    [New LWP 9701]
    [New LWP 9702]
    [Thread debugging using libthread_db enabled]
    Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
    Core was generated by `scfg'.
    Program terminated with signal SIGSEGV, Segmentation fault.
    #0 __memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S:306
    306 ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S: No such file or directory.
    [Current thread is 1 (Thread 0x7f17b0b2a700 (LWP 9701))]
    (gdb) bt
    #0 __memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S:306
    #1 0x00007f17afc77e54 in _IO_new_file_xsputn (f=0x55b439d07da0, data=0x0, n=2) at fileops.c:1327
    #2 0x00007f17afc6d0ab in __GI__IO_fwrite (buf=0x0, size=1, count=2, fp=0x55b439d07da0) at iofwrite.c:39
    #3 0x000055b4377ddb3c in write_node_cfg.part ()
    #4 0x000055b4377a5a39 in save_node_cfg ()
    #5 0x000055b4377c0559 in node_menu ()
    #6 0x000055b4377a4bd8 in main ()

    When I tried it a second time (and again just now) copying "Node 3" to "Node 4", it appeared to not do anything.
    If the settings between node 3 and 4 were the same, then it could appear that way. But if you make a change (say, the phone number) and then copy paste from one node to another, the settings should change.

    Yeah, as I say, node 4's details were blank, and copying from 3 to 4 did not repopulate them (as I thought it would)...
    ...ëîå*

    ... What do you mean? You actually read this Tagline?!?

    ---
    þ Synchronet þ Alterant | an SBBS in Docker on Pi!
  • From Digital Man@VERT to Alterego on Tuesday, September 24, 2019 23:19:42
    Re: Multiserver
    By: Alterego to Digital Man on Wed Sep 25 2019 03:20 pm

    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 09:01 pm

    But what I had tryed, when trying to insert (my earlier message about no INS key), I had discovered CTRL-C and CTRL-V (thinking it might append a new entry) - it infact cause scfg to core dump.
    That's probably how the node's settings got all zeroed out.
    Can you get a backtrace from that core dump? That'd be very helpful in fixing that bug:

    Reading symbols from ../exec/scfg...(no debugging symbols found)...done. [New LWP 9701]
    [New LWP 9702]
    [Thread debugging using libthread_db enabled]
    Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Core was generated by `scfg'.
    Program terminated with signal SIGSEGV, Segmentation fault.
    #0 __memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S:306 306
    ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S: No such file or directory.
    [Current thread is 1 (Thread 0x7f17b0b2a700 (LWP 9701))]
    (gdb) bt
    #0 __memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S:306
    #1 0x00007f17afc77e54 in _IO_new_file_xsputn (f=0x55b439d07da0, data=0x0, n=2) at fileops.c:1327
    #2 0x00007f17afc6d0ab in __GI__IO_fwrite (buf=0x0, size=1, count=2, fp=0x55b439d07da0) at iofwrite.c:39
    #3 0x000055b4377ddb3c in write_node_cfg.part ()
    #4 0x000055b4377a5a39 in save_node_cfg ()
    #5 0x000055b4377c0559 in node_menu ()
    #6 0x000055b4377a4bd8 in main ()

    Can you try again with a *debug* build? LIne numbers (in scfgsave.c) would be very helpful in determining where that crash is actually occurring. It appears to have passed a NULL pointer to fwrite(), but I don't know exactly from where (and none seems obvious looking at the source).

    When I tried it a second time (and again just now) copying "Node 3" to "Node 4", it appeared to not do anything.
    If the settings between node 3 and 4 were the same, then it could appear that way. But if you make a change (say, the phone number) and then copy paste from one node to another, the settings should change.

    Yeah, as I say, node 4's details were blank, and copying from 3 to 4 did not repopulate them (as I thought it would)...

    I noticed that you have to actually open the node config window after pasting for SCFG to recognize that it has changed and it'll then prompt if you want to save those changes. Perhaps that's what you ran into.

    digital man

    Synchronet "Real Fact" #72:
    Synchronet CIOXTRN (created by Deuce) is a 32-bit replacement for DOORWAY. Norco, CA WX: 74.2øF, 50.0% humidity, 0 mph ESE wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • From Digital Man@VERT to Alterego on Tuesday, September 24, 2019 23:23:25
    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 11:19 pm

    Re: Multiserver
    By: Alterego to Digital Man on Wed Sep 25 2019 03:20 pm

    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 09:01 pm

    But what I had tryed, when trying to insert (my earlier message about no INS key), I had discovered CTRL-C and CTRL-V (thinking it might append a new entry) - it infact cause scfg to core dump.
    That's probably how the node's settings got all zeroed out.
    Can you get a backtrace from that core dump? That'd be very helpful in fixing that bug:

    Reading symbols from ../exec/scfg...(no debugging symbols found)...done. [New LWP 9701]
    [New LWP 9702]
    [Thread debugging using libthread_db enabled]
    Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Core was generated by `scfg'. Program terminated with signal SIGSEGV, Segmentation fault.
    #0 __memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S:306 306 ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S: No such file or directory.
    [Current thread is 1 (Thread 0x7f17b0b2a700 (LWP 9701))]
    (gdb) bt
    #0 __memmove_sse2_unaligned_erms () at ../sysdeps/x86_64/multiarch/../multiarch/memmove-vec-unaligned-erms.S:306 #1 0x00007f17afc77e54 in _IO_new_file_xsputn (f=0x55b439d07da0, data=0x0, n=2) at fileops.c:1327
    #2 0x00007f17afc6d0ab in __GI__IO_fwrite (buf=0x0, size=1, count=2, fp=0x55b439d07da0) at iofwrite.c:39
    #3 0x000055b4377ddb3c in write_node_cfg.part ()
    #4 0x000055b4377a5a39 in save_node_cfg ()
    #5 0x000055b4377c0559 in node_menu ()
    #6 0x000055b4377a4bd8 in main ()

    Can you try again with a *debug* build? LIne numbers (in scfgsave.c) would be very helpful in determining where that crash is actually occurring. It appears to have passed a NULL pointer to fwrite(), but I don't know exactly from where (and none seems obvious looking at the source).

    When I tried it a second time (and again just now) copying "Node 3" to "Node 4", it appeared to not do anything.
    If the settings between node 3 and 4 were the same, then it could appear that way. But if you make a change (say, the phone number) and then copy paste from one node to another, the settings should change.

    Yeah, as I say, node 4's details were blank, and copying from 3 to 4 did not repopulate them (as I thought it would)...

    I noticed that you have to actually open the node config window after pasting for SCFG to recognize that it has changed and it'll then prompt if you want to save those changes. Perhaps that's what you ran into.

    Even with that quirk, the pasted changes to my nodes always seem to be saved and are visible in SCFG after exiting and re-running.


    digital man

    This Is Spinal Tap quote #18:
    Sustain, listen to it. Don't hear anything. You would though were it playing. Norco, CA WX: 74.2øF, 50.0% humidity, 0 mph ESE wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • From Alterego@VERT/ALTERANT to Digital Man on Wednesday, September 25, 2019 16:49:12
    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 11:23 pm

    Even with that quirk, the pasted changes to my nodes always seem to be saved and are visible in SCFG after exiting and re-running.

    Hmm... I know I changed - and went into it, and everything was blank (or rather different defaults).

    It also seemed when I tried it that the screen "flashed" as if it was writing an error or something - but it was way too quick :|

    I'll try again though, and see how I go.

    BTW: The "debug" build is made with no RELEASE=1 right? I know I tried that in the past and it gdb still said "no debugging symbols found" - but I'll try again though to be sure.
    ...ëîå*

    ... When two Englishmen meet their first talk is of the weather.

    ---
    þ Synchronet þ Alterant | an SBBS in Docker on Pi!
  • From Alterego@VERT/ALTERANT to Digital Man on Wednesday, September 25, 2019 17:25:45
    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 11:23 pm

    Even with that quirk, the pasted changes to my nodes always seem to be saved and are visible in SCFG after exiting and re-running.

    So I can confirm that that doesnt work for me.

    I created Node 5 - and tried to copy Node 4 to Node 5. I did it many times CTRL-C CTRL-V and it did not copy. However, when I went into node 5 and out again, it prompted me to save.

    As I say, I did it many many times - and there was one time it did do it. So thinking that I must have been dreaming, I blanked out some values and tried again, and could not get it to copy again.

    At one point I CTRL-C copied node 4's entry and pasted CTRL-V on node 4 - and it zapped out the values, all back to some default (that I guess you have internally) - so it must be pasting, but not what I want it to.

    So something is not quite right for me.

    And while I was messing around with this, my BBS on the other node crashed - it "recycled" a few times and then core dumped:

    Core was generated by `sbbs'.
    Program terminated with signal SIGSEGV, Segmentation fault.
    #0 __GI_abort () at abort.c:125
    125 abort.c: No such file or directory.
    [Current thread is 1 (Thread 0x7f6cad5e5700 (LWP 11950))]
    (gdb) bt
    #0 __GI_abort () at abort.c:125
    #1 0x00007f6ce56f2c00 in __libc_message (do_abort=do_abort@entry=2, fmt=fmt@entry=0x7f6ce57e7fd0 "*** Error in `%s': %s: 0x%s ***\n") at ../sysdeps/posix/libc_fatal.c:175
    #2 0x00007f6ce56f8fc6 in malloc_printerr (action=3, str=0x7f6ce57e8098 "double free or corruption (out)", ptr=<optimized out>, ar_ptr=<optimized out>) at malloc.c:5049
    #3 0x00007f6ce56f980e in _int_free (av=0x7f6ce5a1bb00 <main_arena>, p=0x7f6c985e2870, have_lock=0) at malloc.c:3905
    #4 0x00007f6ce707a9ac in msgQueueFree () from /opt/sbbs/exec/libsbbs.so
    #5 0x00007f6ce707aa40 in msgQueueDetach () from /opt/sbbs/exec/libsbbs.so
    #6 0x00007f6ce6e8959f in js_finalize_queue () from /opt/sbbs/exec/libsbbs.so #7 0x00007f6ce51b6341 in JSCompartment::finalizeObjectArenaLists(JSContext*) () from /usr/lib/libmozjs185.so.1.0
    #8 0x00007f6ce51b804f in ?? () from /usr/lib/libmozjs185.so.1.0
    #9 0x00007f6ce51b8787 in ?? () from /usr/lib/libmozjs185.so.1.0
    #10 0x00007f6ce516827a in ?? () from /usr/lib/libmozjs185.so.1.0
    #11 0x00007f6ce6e69f6d in background_thread () from /opt/sbbs/exec/libsbbs.so #12 0x00007f6ce62c54a4 in start_thread (arg=0x7f6cad5e5700) at pthread_create.c:456
    #13 0x00007f6ce576ad0f in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:97

    Anyway, I was trying to get this build to coredump, I rebuilt without RELEASE=1 and it didnt core dump - but I'll leave that debug build running node 4 and use it to make changes to SBBS. Over the past few months that I've been using SBBS, I've had scfg, echocfg and uedit core dump randomly - so when (and if) it does it again, hopefully this build will enable me to give you a better backtrace.
    ...ëîå*

    ... On a clear disk you can seek forever.

    ---
    þ Synchronet þ Alterant | an SBBS in Docker on Pi!
  • From Digital Man@VERT to Alterego on Wednesday, September 25, 2019 00:56:37
    Re: Multiserver
    By: Alterego to Digital Man on Wed Sep 25 2019 04:49 pm

    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 11:23 pm

    Even with that quirk, the pasted changes to my nodes always seem to be saved and are visible in SCFG after exiting and re-running.

    Hmm... I know I changed - and went into it, and everything was blank (or rather different defaults).

    It also seemed when I tried it that the screen "flashed" as if it was writing an error or something - but it was way too quick :|

    Maybe that was when it crashed while writing the node.cnf file.

    I'll try again though, and see how I go.

    BTW: The "debug" build is made with no RELEASE=1 right? I know I tried that in the past and it gdb still said "no debugging symbols found" - but I'll try again though to be sure.

    Yes. You my have to remake your symlinks or copy your executables when switch between release and debug builds too:
    See "Debug builds" at http://wiki.synchro.net/install:nix

    digital man

    Synchronet "Real Fact" #77:
    Rob Swindell still has dozens of BBS-related magazines in his possession. Norco, CA WX: 71.7øF, 59.0% humidity, 0 mph S wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net
  • From Digital Man@VERT to Alterego on Wednesday, September 25, 2019 01:02:14
    Re: Multiserver
    By: Alterego to Digital Man on Wed Sep 25 2019 05:25 pm

    Re: Multiserver
    By: Digital Man to Alterego on Tue Sep 24 2019 11:23 pm

    Even with that quirk, the pasted changes to my nodes always seem to be saved and are visible in SCFG after exiting and re-running.

    So I can confirm that that doesnt work for me.

    I created Node 5 - and tried to copy Node 4 to Node 5. I did it many times CTRL-C CTRL-V and it did not copy. However, when I went into node 5 and out again, it prompted me to save.

    As I say, I did it many many times - and there was one time it did do it. So thinking that I must have been dreaming, I blanked out some values and tried again, and could not get it to copy again.

    At one point I CTRL-C copied node 4's entry and pasted CTRL-V on node 4 - and it zapped out the values, all back to some default (that I guess you have internally) - so it must be pasting, but not what I want it to.

    So something is not quite right for me.

    Well if you could get SCFG to crash while saving the node config again (and provide the backtrace with debug symbols and line numbers) that'll likely point to the cause.

    And while I was messing around with this, my BBS on the other node crashed - it "recycled" a few times and then core dumped:

    Core was generated by `sbbs'.
    Program terminated with signal SIGSEGV, Segmentation fault.

    ...

    #6 0x00007f6ce6e8959f in js_finalize_queue () from /opt/sbbs/exec/libsbbs.so #7 0x00007f6ce51b6341 in JSCompartment::finalizeObjectArenaLists(JSContext*) () from /usr/lib/libmozjs185.so.1.0
    #8 0x00007f6ce51b804f in ?? () from /usr/lib/libmozjs185.so.1.0
    #9 0x00007f6ce51b8787 in ?? () from /usr/lib/libmozjs185.so.1.0
    #10 0x00007f6ce516827a in ?? () from /usr/lib/libmozjs185.so.1.0
    #11 0x00007f6ce6e69f6d in background_thread () from /opt/sbbs/exec/libsbbs.so #12 0x00007f6ce62c54a4 in start_thread (arg=0x7f6cad5e5700) at pthread_create.c:456
    #13 0x00007f6ce576ad0f in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:97

    I'm pretty sure that problem (in js_global.c) has already been fixed in CVS.

    Anyway, I was trying to get this build to coredump, I rebuilt without RELEASE=1 and it didnt core dump - but I'll leave that debug build running node 4 and use it to make changes to SBBS. Over the past few months that I've been using SBBS, I've had scfg, echocfg and uedit core dump randomly - so when (and if) it does it again, hopefully this build will enable me to give you a better backtrace. ...ëîå*

    Thanks. Nothing is random, but it can be difficult to reproduce issues sometimes. Memory corruptions are like ricochetting bullets, you never know if/when they'll cause damage and usually... they don't.

    digital man

    Synchronet/BBS Terminology Definition #64:
    SMB = Synchronet Message Base (e.g. smblib)
    Norco, CA WX: 71.7øF, 59.0% humidity, 0 mph S wind, 0.00 inches rain/24hrs

    ---
    þ Synchronet þ Vertrauen þ Home of Synchronet þ [vert/cvs/bbs].synchro.net