Hey guys,
I noticed I couldn't ssh into my bbs today (I prefer to not use telnet basically ever) but telnet was working fine. I ssh'd into the server itself and noticed in 'top' that mis was eating up 100% of my CPU. I bounced the mis service and now it's hovering around 5ish percent of
CPU. There's nothing in error.log and nothing interesting in mis.log.
Any ideas?
itself and noticed in 'top' that mis was eating up 100% of my CPU. I bounced the mis service and now it's hovering around 5ish percent of
CPU. There's nothing in error.log and nothing interesting in mis.log.
Any ideas?
itself and noticed in 'top' that mis was eating up 100% of my CPU. I bounced the mis service and now it's hovering around 5ish percent of CPU. There's nothing in error.log and nothing interesting in mis.log. Any ideas?
If this is an ongoing problem you could consider upgrading to the latest test versions and we can go from there if the problem persists.
Were any users "stuck" online?
itself and noticed in 'top' that mis was eating up 100% of my CP bounced the mis service and now it's hovering around 5ish percen CPU. There's nothing in error.log and nothing interesting in mis Any ideas?
If this is an ongoing problem you could consider upgrading to the lat test versions and we can go from there if the problem persists.
Were any users "stuck" online?
Hi g00r00,
Yeah, it's back up to 100% again for some reason, though I'm able to use the BBS for now. Nothing in any mystic logs or in syslog, for that
matter.
No users stuck online. I'm on alpha 36, 32 bit linux. So I should
upgrade, huh? I can give that a shot.
Curious you are only loading MIS now with A37 I think the late A36. g00r00 changed the name of mis2 to mis and removed mis v1 all together.
I ask this due to the possibility of you using both mis/mis2 still and
the events had been transfered to the new mis so only one instance was needed. :)
Hi Pequito, thanks :) I actually completely killed the vm slice where the mis/mis2 configuration issues lived. I then moved to install a fresh
a36. I just now upgraded to a37, and I'm keeping an eye open for the
.bsy file issue I ran into before. Fingers crossed...
There will also be .mem files created so if you have a crash it can (hopefully) tell me an estimated line in the code where it happened.
I'll also add in any additional logging I can surrounding any issue
you're having if needed and I can update the prealpha site instantly.
The downsides with the latest prealphas are that I accidentally compiled it with BINKP debug logging so those BINKP logs get pretty cluttered.
The binaries are also up to 10 times the size they normally are, and things in general will run a little slower... but it should go a long
way to help me resolve your issues.
One thing that was weird was that a file, /mystic/logs/eventsignal.txt
was being created and filling up faster than I could kill it. It was filling up like hundreds of gigs a day. My only workaround thus far was
to actually shut down mis, kill eventsignal.txt, create an empty eventsignal.txt, and chmod -w the file (so it couldn't be written to).
However now I do keep running into the issue of mis taking up 100% of
CPU. Any ideas where to start troubleshooting?
One thing that was weird was that a file, /mystic/logs/eventsignal.txt was being created and filling up faster than I could kill it. It was filling up like hundreds of gigs a day. My only workaround thus far was to actually shut down mis, kill eventsignal.txt, create an empty eventsignal.txt, and chmod -w the file (so it couldn't be written to).
Any idea of what events were in it that was filling it up?
However now I do keep running into the issue of mis taking up 100% of CPU. Any ideas where to start troubleshooting?
I am not able to reproduce any issue like this so far, but I haven't given it >enough time because the past couple of days have been with friends and family >for the most part. The little I had for Mystic I've been working on things >easy to pick up and put down like the text/ANSI editors (which are almost >done).
If you know what event(s) were filling up the debug event file that might be a >clue.
Someone else was suggesting that it might be related to SSH - do you think that
could be possible? Are you able to disable SSH to see if it stops the problem?
Does it happen immediately every time or just randomly after running for a >while?
I'll try to see if I can get it to happen soon. I am going to put my BBS up >via SSH on Linux/64 for a while and work on any stability issues.
--- Mystic BBS v1.12 A37 2017/12/26 (Windows/32)
* Origin: Sector 7 [Mystic BBS WHQ] (21:1/108)
g00r00 wrote to esc <=-
One thing that was weird was that a file, /mystic/logs/eventsignal.txt
was being created and filling up faster than I could kill it. It was filling up like hundreds of gigs a day. My only workaround thus far was
to actually shut down mis, kill eventsignal.txt, create an empty eventsignal.txt, and chmod -w the file (so it couldn't be written to).
Any idea of what events were in it that was filling it up?
Nugax wrote to All <=-
Might want to set a file size limit on log files too... just a thought
Any idea of what events were in it that was filling it up?
I am not able to reproduce any issue like this so far, but I haven't
given it enough time because the past couple of days have been with friends and family for the most part. The little I had for Mystic I've been working on things easy to pick up and put down like the text/ANSI editors (which are almost done).
If you know what event(s) were filling up the debug event file that
might be a clue.
Someone else was suggesting that it might be related to SSH - do you
think that could be possible? Are you able to disable SSH to see if it stops the problem? Does it happen immediately every time or just
randomly after running for a while?
esc wrote to g00r00 <=-
I _could_ disable ssh and use telnet but not for a couple of weeks. I'm camping at various RV parks with crap open guest wifi and am not a huge fan of telnetting in the clear. Sadly, wrapping stuff in my vpn is
tough, because it totally crushes network perf.
Any idea of what events were in it that was filling it up?
Sadly, no. I suppose hypothetically I could remove the temporary
non-write file I created and see...
I am not able to reproduce any issue like this so far, but I haven't given it enough time because the past couple of days have been with friends and family for the most part. The little I had for Mystic I' been working on things easy to pick up and put down like the text/ANS editors (which are almost done).
Dude! I'm not even paying you. You owe me nothing. Please don't ever
feel the need to explain :)
It seems to take a good day or so for this to happen. It's pretty
bizarre. I wonder if I'm being portscanned or someone's able to ddos mis somehow. Interestingly enough, when it's using 100% cpu, ssh doesn't
work at all, but telnet still does.
I _could_ disable ssh and use telnet but not for a couple of weeks. I'm camping at various RV parks with crap open guest wifi and am not a huge fan of telnetting in the clear. Sadly, wrapping stuff in my vpn is
tough, because it totally crushes network perf.
If you have a Linux box on your etwork, why not use SSH forwarding to securely carry a telnet session over the Internet? I do this on my Windows XP laptop that won't run SyncTerm. I run PuTTY with the port forwarding setup to the BBS. That PuTTY session connects to a Linux box inside the network. Once the SSH session is active, I point Netrunner
at localhost on the appropriate port for the session ! want. Bingo!
all BBS communication is inside SSH encryption. :)
Think he meant check mystic/cfg/events and see what is active/running there. If you run in daemon maybe try out in server mode for a day so could also monitor it and see what happens? :)
Other thing I can think of close out mis and in the cfg re-create the telnet server and restart mis and see if this might help? Also when you upgraded from a?? to a37 did you run the upgrade at least once? :)
esc wrote to vk3jed <=-
Yeah, I suppose I could port forward from putty. I haven't yet switched off ssh in mis, cpu usage is still hovering under 5% :)
| Sysop: | sneaky |
|---|---|
| Location: | Ashburton,NZ |
| Users: | 2 |
| Nodes: | 8 (0 / 8) |
| Uptime: | 82:18:31 |
| Calls: | 2,120 |
| Calls today: | 1 |
| Files: | 11,149 |
| D/L today: |
30 files (10,780K bytes) |
| Messages: | 950,616 |