MUD dies after running for hours

[rgautier]

Two days in a row now, the MUD process has just ended around 11:45 PM for no particular reason. No one is logged into it, the vme process just ends (no error logged, no input logged), no process error on stderr.

Last logged events are merely zone file errors that had occurred throughout the day regarding objects missing rooms.

[prool]

me too ;)

gdb is founded crashes in lines:

dilrun.cpp:1070 event.cpp:222 event.cpp:227 event.cpp:231

[rgautier]

This is the tail of last night's vme.log

Fri Jul 24 21:20:49 2020 :: hobgob/Zone Reset Error: No west direction from room hidden_cave in door. Fri Jul 24 21:20:49 2020 :: hobgob/Reset Error: Don't know where to put pseudo@hobgob Fri Jul 24 21:20:49 2020 :: hobgob/Reset Error: Don't know where to put lover_2@hobgob Fri Jul 24 21:20:49 2020 :: hobgob/Reset Error: Don't know where to put kitten@hobgob Fri Jul 24 21:20:49 2020 :: hobgob/Reset Error: Don't know where to put mikisiw@hobgob Fri Jul 24 21:23:46 2020 :: sorcerer/Reset Error: Don't know where to put good_mage@sorcerer Fri Jul 24 21:25:36 2020 :: Kicking out idle player waiting for name. Fri Jul 24 21:30:36 2020 :: Kicking out idle player waiting for name. Fri Jul 24 21:31:33 2020 :: assassin/Reset Error: Don't know where to put assassin_guard@assassin Fri Jul 24 21:35:36 2020 :: Kicking out idle player waiting for name. Fri Jul 24 21:39:27 2020 :: offices/Reset Error: Don't know where to put shredder@offices Fri Jul 24 21:39:27 2020 :: offices/Reset Error: Don't know where to put shredder@offices Fri Jul 24 21:39:41 2020 :: midgaard/Zone Reset Error: No down direction from room wall_rd2 in door. Fri Jul 24 21:39:41 2020 :: midgaard/Zone Reset Error: No down direction from room wall_rd11 in door. Fri Jul 24 21:40:36 2020 :: Kicking out idle player waiting for name. Fri Jul 24 21:40:52 2020 :: halfzon/Reset Error: Don't know where to put lasrarner@halfzon Fri Jul 24 21:40:52 2020 :: halfzon/Reset Error: Don't know where to put hanook@halfzon Fri Jul 24 21:40:52 2020 :: halfzon/Reset Error: Don't know where to put pallowaz@halfzon Fri Jul 24 21:42:01 2020 :: hobgob/Zone Reset Error: No west direction from room hidden_cave in door. Fri Jul 24 21:42:01 2020 :: hobgob/Reset Error: Don't know where to put pseudo@hobgob Fri Jul 24 21:42:01 2020 :: hobgob/Reset Error: Don't know where to put lover_2@hobgob Fri Jul 24 21:42:01 2020 :: hobgob/Reset Error: Don't know where to put kitten@hobgob Fri Jul 24 21:42:01 2020 :: hobgob/Reset Error: Don't know where to put mikisiw@hobgob Fri Jul 24 21:45:21 2020 :: sorcerer/Reset Error: Don't know where to put good_mage@sorcerer Fri Jul 24 21:45:36 2020 :: Kicking out idle player waiting for name. Fri Jul 24 21:45:52 2020 :: Error: patrol_guard1@midgaard had heartbeat of 1. Fri Jul 24 21:45:52 2020 :: Error: patrol_guard1@midgaard had heartbeat of 1. Fri Jul 24 21:45:52 2020 :: DIL heartbeat was

[rgautier]

Hrm - just got this: [1]- Segmentation fault nohup ../bin/vme 2> vmeerr.log > nohupvme.log [3]+ Exit 1 nohup ../bin/mplex -w -p 4280 2> mplexerr.log > nohupmplex.log

vme.log on attempted restart - Sat Jul 25 07:37:27 2020 :: haon_dor/DIL walk_room() unable to locate room throne@dwarves Sat Jul 25 07:37:27 2020 :: haon_dor/DIL walk_room() unable to locate room throne@dwarves Sat Jul 25 07:37:36 2020 :: midgaard/DIL watercurrent(): No such direction 1 in room river1@midgaard Sat Jul 25 07:37:41 2020 :: Booting Eliza Sat Jul 25 07:37:41 2020 :: Booting Eliza Done. Sat Jul 25 07:37:51 2020 :: A multi-host has connected to the game. Sat Jul 25 07:37:51 2020 :: Sending Mudname Valhalla Test Mud, Len 18 Sat Jul 25 07:37:51 2020 :: Sending Default Colors to multihost. Sat Jul 25 07:39:19 2020 :: goblin_cave/DIL walk_room() unable to locate room pt_d_fo@elfdom Sat Jul 25 07:39:44 2020 :: goblin_cave/DIL walk_room() unable to locate room fireplac@elfdom

[Seifert69]

I'll see if I can figure out what it is. Hang on 24 hours. :)

On Sat, Jul 25, 2020 at 1:40 PM Serge notifications@github.com wrote:

me too ;)

gdb is founded crashes in lines:

dilrun.cpp:1070 event.cpp:222 event.cpp:227 event.cpp:231

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663845693, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXWHGK7AH6HIUJ2VLXDR5LADLANCNFSM4PHNIISA .

[Seifert69]

Lines numbers dont match up precisely with my code.

What OS'es are you running on?

Can you possibly mail me a backtrace from gdb?

And a few of the key frames?

If it's in the line I think it is, then I'd love to see a print of:

if (((dilprg )((unit_fptr )tmp_event->arg2)->data)->fp->tmpl->prgname)

I'm thinking maybe "fp" or "tmpl" might be NULL ?

And just to be sure: You're not compiling zones in the zone/ dir while the MUD is running, right?

Best, M

On Sat, Jul 25, 2020 at 1:40 PM Serge notifications@github.com wrote:

me too ;)

gdb is founded crashes in lines:

dilrun.cpp:1070 event.cpp:222 event.cpp:227 event.cpp:231

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663845693, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXWHGK7AH6HIUJ2VLXDR5LADLANCNFSM4PHNIISA .

[prool]

Lines numbers dont match up precisely with my code.

Maybe, I'm using old version of code.

I'm retesting new MUD code under gdb, wait plz about 24h.

My OS is Ubuntu 18.04.4 x86_64

[Seifert69]

Ok.

I'm running for 14 days without issues. Please get the latest. I'm updating a lot these days

On Sat, Jul 25, 2020, 18:47 Serge notifications@github.com wrote:

Lines numbers dont match up precisely with my code.

Maybe, I'm using old version of code.

I'm retesting new MUD code under gdb, wait plz about 24h.

My OS is Ubuntu 18.04.4 x86_64

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663876128, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXRHP44HDOAKJJWFRY3R5MEDPANCNFSM4PHNIISA .

[rgautier]

Am running CentOS 7, patched. No core dumps.. (I might have them turned off?)

I was asleep when it crashed, so not compiling zones.

I cloned it from the 20th....did you update since then? I just ran a git pull, and only Makefile has changed (I assume for the C++ std line.)

On Sat, Jul 25, 2020, 12:57 PM Michael Seifert notifications@github.com wrote:

Ok.

I'm running for 14 days without issues. Please get the latest. I'm updating a lot these days

On Sat, Jul 25, 2020, 18:47 Serge notifications@github.com wrote:

Lines numbers dont match up precisely with my code.

Maybe, I'm using old version of code.

I'm retesting new MUD code under gdb, wait plz about 24h.

My OS is Ubuntu 18.04.4 x86_64

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663876128 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AOOOMXRHP44HDOAKJJWFRY3R5MEDPANCNFSM4PHNIISA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663877112, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFPYGDH7K2DELFXMVJFJ5DR5MFHHANCNFSM4PHNIISA .

[Seifert69]

I figured out a neat way to reproduce (remove usleep() delay to accelerate time by 500x+). The bug is super complicated and will take a while to resolve. I think I'll end up having to rewrite a few things to classes to make it more bullet proof. Will let you know how it goes. PS. The production MUD can run for 20+ days without issues, it's apparently only the combo on the scaled back version.

On Sat, Jul 25, 2020 at 9:20 PM Rich Gautier notifications@github.com wrote:

Am running CentOS 7, patched. No core dumps.. (I might have them turned off?)

I was asleep when it crashed, so not compiling zones.

On Sat, Jul 25, 2020, 12:57 PM Michael Seifert notifications@github.com wrote:

Ok.

I'm running for 14 days without issues. Please get the latest. I'm updating a lot these days

On Sat, Jul 25, 2020, 18:47 Serge notifications@github.com wrote:

Lines numbers dont match up precisely with my code.

Maybe, I'm using old version of code.

I'm retesting new MUD code under gdb, wait plz about 24h.

My OS is Ubuntu 18.04.4 x86_64

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663876128 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AOOOMXRHP44HDOAKJJWFRY3R5MEDPANCNFSM4PHNIISA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663877112 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AAFPYGDH7K2DELFXMVJFJ5DR5MFHHANCNFSM4PHNIISA

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663891290, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXXUDG6VPTYFK7TPCO3R5MV75ANCNFSM4PHNIISA .

[rgautier]

I believe that one thing that made it crash after a time was leaving a browser connected to it, without it logging in. The MUD would disconnect it, over and over. Last night, I didn't leave my browser open to it, and it didn't crash...has been up for 36 hours now.

On Sun, Jul 26, 2020 at 2:24 PM Michael Seifert notifications@github.com wrote:

I figured out a neat way to reproduce (remove usleep() delay to accelerate time by 500x+). The bug is super complicated and will take a while to resolve. I think I'll end up having to rewrite a few things to classes to make it more bullet proof. Will let you know how it goes. PS. The production MUD can run for 20+ days without issues, it's apparently only the combo on the scaled back version.

On Sat, Jul 25, 2020 at 9:20 PM Rich Gautier notifications@github.com wrote:

Am running CentOS 7, patched. No core dumps.. (I might have them turned off?)

I was asleep when it crashed, so not compiling zones.

On Sat, Jul 25, 2020, 12:57 PM Michael Seifert <notifications@github.com

wrote:

Ok.

I'm running for 14 days without issues. Please get the latest. I'm updating a lot these days

On Sat, Jul 25, 2020, 18:47 Serge notifications@github.com wrote:

Lines numbers dont match up precisely with my code.

Maybe, I'm using old version of code.

I'm retesting new MUD code under gdb, wait plz about 24h.

My OS is Ubuntu 18.04.4 x86_64

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663876128 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AOOOMXRHP44HDOAKJJWFRY3R5MEDPANCNFSM4PHNIISA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663877112 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AAFPYGDH7K2DELFXMVJFJ5DR5MFHHANCNFSM4PHNIISA

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663891290 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AOOOMXXUDG6VPTYFK7TPCO3R5MV75ANCNFSM4PHNIISA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-664023509, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFPYGFFKO36M5P7UQJKT23R5RYFBANCNFSM4PHNIISA .

[Seifert69]

I can also make it crash without now :-)

I think maybe some of my newer DIL code is triggering very rarely.

I'm running valgrind now, found a small handful of issues that I fixed but I can still crash it.

Running prolonged now in valgrind (because it slows down the machine). It might take a few weeks to figure it out. It's one of those nightmare bugs :))

Best

On Mon, Jul 27, 2020, 02:07 Rich Gautier notifications@github.com wrote:

I believe that one thing that made it crash after a time was leaving a browser connected to it, without it logging in. The MUD would disconnect it, over and over. Last night, I didn't leave my browser open to it, and it didn't crash...has been up for 36 hours now.

On Sun, Jul 26, 2020 at 2:24 PM Michael Seifert notifications@github.com wrote:

I figured out a neat way to reproduce (remove usleep() delay to accelerate time by 500x+). The bug is super complicated and will take a while to resolve. I think I'll end up having to rewrite a few things to classes to make it more bullet proof. Will let you know how it goes. PS. The production MUD can run for 20+ days without issues, it's apparently only the combo on the scaled back version.

On Sat, Jul 25, 2020 at 9:20 PM Rich Gautier notifications@github.com wrote:

Am running CentOS 7, patched. No core dumps.. (I might have them turned off?)

I was asleep when it crashed, so not compiling zones.

On Sat, Jul 25, 2020, 12:57 PM Michael Seifert < notifications@github.com

wrote:

Ok.

I'm running for 14 days without issues. Please get the latest. I'm updating a lot these days

On Sat, Jul 25, 2020, 18:47 Serge notifications@github.com wrote:

Lines numbers dont match up precisely with my code.

Maybe, I'm using old version of code.

I'm retesting new MUD code under gdb, wait plz about 24h.

My OS is Ubuntu 18.04.4 x86_64

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663876128 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AOOOMXRHP44HDOAKJJWFRY3R5MEDPANCNFSM4PHNIISA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663877112 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AAFPYGDH7K2DELFXMVJFJ5DR5MFHHANCNFSM4PHNIISA

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-663891290 , or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AOOOMXXUDG6VPTYFK7TPCO3R5MV75ANCNFSM4PHNIISA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-664023509 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AAFPYGFFKO36M5P7UQJKT23R5RYFBANCNFSM4PHNIISA

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-664058831, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXQPFENN6GT55WW5JOTR5TAKRANCNFSM4PHNIISA .

[rgautier]

So, are you saying you're only releasing the scaled back version? :) I definitely know we're missing some zones (all those attempted trips to elfdom by my NPCs). Have you thought of only including a simple Zone with no actions that tie into missing zones for the release version?

[Seifert69]

We released everything except zones special to Valhalla MUD. Those zones are not really mine to give away :o) If there are some particular zones you would like to see here, drop me an email and I'll poke Ken and see what he says. I like your idea though. :o)

On Mon, Jul 27, 2020 at 12:54 PM Rich Gautier notifications@github.com wrote:

So, are you saying you're only releasing the scaled back version? :) I definitely know we're missing some zones (all those attempted trips to elfdom by my NPCs). Have you thought of only including a simple Zone with no actions that tie into missing zones for the release version?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-664319779, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXWUHMKFVC6WIRSEDDTR5VMHNANCNFSM4PHNIISA .

[Seifert69]

I've located and fixed the bug - which by the way was extremely hard :-) The issue happened if a run_dil() had recursively called itself, then there was a risk that it would free the dilprg struct twice.

Also ran everything through valgrind and caught surprisingly few additional errors.

Will upload sources soon.

On Mon, Jul 27, 2020 at 3:54 PM Michael Seifert michael@corleone.com wrote:

We released everything except zones special to Valhalla MUD. Those zones are not really mine to give away :o) If there are some particular zones you would like to see here, drop me an email and I'll poke Ken and see what he says. I like your idea though. :o)

On Mon, Jul 27, 2020 at 12:54 PM Rich Gautier notifications@github.com wrote:

So, are you saying you're only releasing the scaled back version? :) I definitely know we're missing some zones (all those attempted trips to elfdom by my NPCs). Have you thought of only including a simple Zone with no actions that tie into missing zones for the release version?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-664319779, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXWUHMKFVC6WIRSEDDTR5VMHNANCNFSM4PHNIISA .

[rgautier]

Pulled a fresh and running now - will let you know if it dies in the next 48 hours for no reason. But I do see you've done a lot of zone cleanup work...nice job!

[rgautier]

Crashed at about 1PM today. No one was logged into it. tail of vme.log is below - doesn't look like a nesting issue....

Is there anything I can do to get to the bottom of this? (I'm open to trying anything you need to help debug it)

Wed Jul 29 12:45:24 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:45:57 2020 :: run_dil() officerprg has nest 2 Wed Jul 29 12:46:35 2020 :: patrol guard was given 2 copper pieces by DIL guard3@midwalls. Wed Jul 29 12:46:37 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:46:37 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:47:29 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:47:29 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:48:22 2020 :: gate guard was given 2 copper pieces by DIL gate_n1@midwalls. Wed Jul 29 12:48:26 2020 :: patrol guard was given 2 copper pieces by DIL guard8@midwalls. Wed Jul 29 12:49:04 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:49:04 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:49:21 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:51:06 2020 :: gate guard was given 2 copper pieces by DIL gate_e2@midwalls. Wed Jul 29 12:51:19 2020 :: offices/Reset Error: Don't know where to put shredder@offices Wed Jul 29 12:51:19 2020 :: offices/Reset Error: Don't know where to put shredder@offices Wed Jul 29 12:51:26 2020 :: patrol guard was given 2 copper pieces by DIL guard11@midwalls. Wed Jul 29 12:52:20 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:52:20 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:55:53 2020 :: patrol guard was given 2 copper pieces by DIL guard7@midwalls. Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put lasrarner@halfzon Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put hanook@halfzon Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put pallowaz@halfzon Wed Jul 29 12:56:50 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:56:50 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:57:11 2020 :: kingc/Reset Error: Don't know where to put pool_monster@kingc

mplex.log from same time frame: Wed Jul 29 12:57:36 2020 :: Ping received Wed Jul 29 12:57:40 2020 :: Read to queue: EOF on socket read (eno 11). Wed Jul 29 12:57:40 2020 :: Protocol: parse_incoming error. Wed Jul 29 12:57:40 2020 :: ERROR READING FROM MUD. Wed Jul 29 12:57:41 2020 :: Open connection to server on 127.0.0.1 4999. Wed Jul 29 12:57:41 2020 :: No connect allowed to server. Wed Jul 29 12:57:42 2020 :: Unable to connect to MUD server 127.0.0.1:4999 Wed Jul 29 12:57:42 2020 :: Waiting for a MUDHook to get reconnected... Wed Jul 29 12:57:43 2020 :: Open connection to server on 127.0.0.1 4999. Wed Jul 29 12:57:43 2020 :: No connect allowed to server.

[Seifert69]

Oh no. :-( I'll look into it again :-)

On Wed, Jul 29, 2020, 23:06 Rich Gautier notifications@github.com wrote:

Crashed at about 1PM today. No one was logged into it. tail of vme.log is below - doesn't look like a nesting issue....

Is there anything I can do to get to the bottom of this? (I'm open to trying anything you need to help debug it)

Wed Jul 29 12:45:24 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:45:57 2020 :: run_dil() officerprg has nest 2 Wed Jul 29 12:46:35 2020 :: patrol guard was given 2 copper pieces by DIL guard3@midwalls. Wed Jul 29 12:46:37 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:46:37 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:47:29 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:47:29 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:48:22 2020 :: gate guard was given 2 copper pieces by DIL gate_n1@midwalls. Wed Jul 29 12:48:26 2020 :: patrol guard was given 2 copper pieces by DIL guard8@midwalls. Wed Jul 29 12:49:04 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:49:04 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:49:21 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:51:06 2020 :: gate guard was given 2 copper pieces by DIL gate_e2@midwalls. Wed Jul 29 12:51:19 2020 :: offices/Reset Error: Don't know where to put shredder@offices Wed Jul 29 12:51:19 2020 :: offices/Reset Error: Don't know where to put shredder@offices Wed Jul 29 12:51:26 2020 :: patrol guard was given 2 copper pieces by DIL guard11@midwalls. Wed Jul 29 12:52:20 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:52:20 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:55:53 2020 :: patrol guard was given 2 copper pieces by DIL guard7@midwalls. Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put lasrarner@halfzon Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put hanook@halfzon Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put pallowaz@halfzon Wed Jul 29 12:56:50 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:56:50 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:57:11 2020 :: kingc/Reset Error: Don't know where to put pool_monster@kingc

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-665927762, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXUBGBD7LWIOVWAKMBLR6CFNJANCNFSM4PHNIISA .

[Seifert69]

No luck yet in reproducing it. Can you double check you're running on the new version. If you can run the server from gdb and let me know how it terminates, either the stacktrace or the exit code please. Please do pull the latest source as I differentiated the exit codes (they all tended to exit with 1 and were undistinguishable) . Thanks :)

On Wed, Jul 29, 2020 at 11:06 PM Rich Gautier notifications@github.com wrote:

Crashed at about 1PM today. No one was logged into it. tail of vme.log is below - doesn't look like a nesting issue....

Is there anything I can do to get to the bottom of this? (I'm open to trying anything you need to help debug it)

Wed Jul 29 12:45:24 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:45:57 2020 :: run_dil() officerprg has nest 2 Wed Jul 29 12:46:35 2020 :: patrol guard was given 2 copper pieces by DIL guard3@midwalls. Wed Jul 29 12:46:37 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:46:37 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:47:29 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:47:29 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:48:22 2020 :: gate guard was given 2 copper pieces by DIL gate_n1@midwalls. Wed Jul 29 12:48:26 2020 :: patrol guard was given 2 copper pieces by DIL guard8@midwalls. Wed Jul 29 12:49:04 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:49:04 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:49:21 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:51:06 2020 :: gate guard was given 2 copper pieces by DIL gate_e2@midwalls. Wed Jul 29 12:51:19 2020 :: offices/Reset Error: Don't know where to put shredder@offices Wed Jul 29 12:51:19 2020 :: offices/Reset Error: Don't know where to put shredder@offices Wed Jul 29 12:51:26 2020 :: patrol guard was given 2 copper pieces by DIL guard11@midwalls. Wed Jul 29 12:52:20 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:52:20 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:55:53 2020 :: patrol guard was given 2 copper pieces by DIL guard7@midwalls. Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put lasrarner@halfzon Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put hanook@halfzon Wed Jul 29 12:56:45 2020 :: halfzon/Reset Error: Don't know where to put pallowaz@halfzon Wed Jul 29 12:56:50 2020 :: run_dil() do_say has nest 2 Wed Jul 29 12:56:50 2020 :: run_dil() pain_dil has nest 2 Wed Jul 29 12:57:11 2020 :: kingc/Reset Error: Don't know where to put pool_monster@kingc

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-665927762, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXUBGBD7LWIOVWAKMBLR6CFNJANCNFSM4PHNIISA .

[rgautier]

I had pulled and 'make clean' / 'make all' before running last night. The issue with running from gdb is I'm disconnecting from my cloud server - so to keep it running, I'm running in nohup and background... Not sure how to interactively run in gdb and then disconnect/background the task.

Suggestions?

I've redirected stderr so I can capture it in an independent log. Hopefully I'll capture the exit code when/if it crashes tomorrow...fingers crossed.

[Seifert69]

Hi Rich,

Thank you. I can't make mine "crash" in any way. However, if I run it for 21 day, it will get killed by Linux for using too much memory.

I can see there's a memory leak related to DIL. I'll start working on that but it'll take a few weeks I suspect. My plan is to change the diltemplate and dilframe structs to classes, to maintain counters, and thereby figure out where the memory is lost.

If you look in:

less /var/log/kern.log

You might be able to see if Linux killed your vme process.

Best, Michael

On Thu, Jul 30, 2020, 00:48 Rich Gautier notifications@github.com wrote:

I had pulled and 'make clean' / 'make all' before running last night. The issue with running from gdb is I'm disconnecting from my cloud server - so to keep it running, I'm running in nohup and background... Not sure how to interactively run in gdb and then disconnect/background the task.

Suggestions?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-665969608, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXRKA65QUDDXAXZ3DD3R6CRMVANCNFSM4PHNIISA .

[prool]

I confirm the leak. Look at the graph http://prooltest-4g.kharkov.org/munin/localdomain/localhost.localdomain/memory.html Leak from 9:00 Wed to current time. Towers is the Minecraft game of my grandson ;-)

[Seifert69]

Thank you. Think I found it but it'll take a little while to fix. Not as simple as tracking with classes. There's some really odd code around dilframe handling and reallocate.

On Thu, Jul 30, 2020, 11:22 Serge notifications@github.com wrote:

I confirm the leak. Look at the graph

http://prooltest-4g.kharkov.org/munin/localdomain/localhost.localdomain/memory.html Leak from 9:00 Wed to current time. Towers is the Minecraft game of my grandson ;-)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-666253965, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXST6BU2XVN35EH7EOTR6E3U5ANCNFSM4PHNIISA .

[Seifert69]

If you have the time please try to pull and run again.

There's still a leak, looks like it's in DIL variables. But I suspect you can run for long now without running out of resources.

I'll look at it but this time it'll take a few weeks unless it still goes down after 11 hours :)))

On Thu, Jul 30, 2020 at 12:48 AM Rich Gautier notifications@github.com wrote:

I had pulled and 'make clean' / 'make all' before running last night. The issue with running from gdb is I'm disconnecting from my cloud server - so to keep it running, I'm running in nohup and background... Not sure how to interactively run in gdb and then disconnect/background the task.

Suggestions?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-665969608, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXRKA65QUDDXAXZ3DD3R6CRMVANCNFSM4PHNIISA .

[rgautier]

Cannot compile on new pull:

g++ -I. -I.. -I vmc -DVMC -DLINUX -DPOSIX -pipe -m64 -std=c++11 -DMUD_DEBUG -g -Wall -O0 ../../build//vmc/vmc.o ../../build//vmc/vmc_process.o ../../build//v mc/utility.o ../../build//vmc/structs.o ../../build//vmc/destruct.o ../../build/ /vmc/spec_assign.o ../../build//vmc/money.o ../../build//vmc/pp1.o ../../build// vmc/pp2.o ../../build//vmc/pp3.o ../../build//vmc/pp4.o ../../build//vmc/pp5.o . ./../build//vmc/pp6.o ../../build//vmc/pp7.o ../../build//vmc/pp8.o ../../build/ /vmc/vmc_db_file.o ../../build//vmc/tmp_vmcpar.o ../../build//vmc/tmp_vmclex.o . ./../build//vmc/tmp_dilpar.o ../../build//vmc/tmp_dillex.o ../../build//color.o ../../build//textutil.o ../../build//namelist.o ../../build//files.o ../../build //extra.o ../../build//bytestring.o ../../build//common.o ../../build//intlist.o ../../build//dilshare.o -lfl -g -o ../../bin/vmc ../../build//dilshare.o: In function dilprg::dilprg(unit_data*, int)': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:84: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:85: undefined reference to dil_list' ../../build//dilshare.o: In functiondilprg::~dilprg()': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference tog_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:132: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:134: undefined reference todil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference todil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:137: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:138: undefined reference todil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:143: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:147: undefined reference todil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:148: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:175: undefined reference todil_free_template(diltemplate*, int)'

[Seifert69]

Blast it, I keep forgetting the Makefile, sorry. It's up now :)

On Fri, Jul 31, 2020 at 1:02 AM Rich Gautier notifications@github.com wrote:

Cannot compile on new pull:

g++ -I. -I.. -I vmc -DVMC -DLINUX -DPOSIX -pipe -m64 -std=c++11 -DMUD_DEBUG -g -Wall -O0 ../../build//vmc/vmc.o ../../build//vmc/vmc_process.o ../../build//v mc/utility.o ../../build//vmc/structs.o ../../build//vmc/destruct.o ../../build/ /vmc/spec_assign.o ../../build//vmc/money.o ../../build//vmc/pp1.o ../../build// vmc/pp2.o ../../build//vmc/pp3.o ../../build//vmc/pp4.o ../../build//vmc/pp5.o . ./../build//vmc/pp6.o ../../build//vmc/pp7.o ../../build//vmc/pp8.o ../../build/ /vmc/vmc_db_file.o ../../build//vmc/tmp_vmcpar.o ../../build//vmc/tmp_vmclex.o . ./../build//vmc/tmp_dilpar.o ../../build//vmc/tmp_dillex.o ../../build//color.o ../../build//textutil.o ../../build//namelist.o ../../build//files.o ../../build //extra.o ../../build//bytestring.o ../../build//common.o ../../build//intlist.o ../../build//dilshare.o -lfl -g -o ../../bin/vmc ../../build//dilshare.o: In function dilprg::dilprg(unit_data, int)': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:84: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:85: undefined reference to dil_list' ../../build//dilshare.o: In function dilprg::~dilprg()': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:132: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:134: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:137: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:138: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:143: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:147: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:148: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:175: undefined reference to dil_free_template(diltemplate, int)'

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-666763017, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXTEYKPULFTZTHLCOWTR6H3XZANCNFSM4PHNIISA .

[Seifert69]

So I found the memory leak. stack.init(10) apparently was the culprit. I've never seen that class before.

Anyway there seems to be one pesky, elusive bug I'm still hunting for. I'll let you know when I've fixed it, then it should be all good.

Best,

On Fri, Jul 31, 2020 at 8:03 AM Michael Seifert michael@corleone.com wrote:

Blast it, I keep forgetting the Makefile, sorry. It's up now :)

On Fri, Jul 31, 2020 at 1:02 AM Rich Gautier notifications@github.com wrote:

Cannot compile on new pull:

g++ -I. -I.. -I vmc -DVMC -DLINUX -DPOSIX -pipe -m64 -std=c++11 -DMUD_DEBUG -g -Wall -O0 ../../build//vmc/vmc.o ../../build//vmc/vmc_process.o ../../build//v mc/utility.o ../../build//vmc/structs.o ../../build//vmc/destruct.o ../../build/ /vmc/spec_assign.o ../../build//vmc/money.o ../../build//vmc/pp1.o ../../build// vmc/pp2.o ../../build//vmc/pp3.o ../../build//vmc/pp4.o ../../build//vmc/pp5.o . ./../build//vmc/pp6.o ../../build//vmc/pp7.o ../../build//vmc/pp8.o ../../build/ /vmc/vmc_db_file.o ../../build//vmc/tmp_vmcpar.o ../../build//vmc/tmp_vmclex.o . ./../build//vmc/tmp_dilpar.o ../../build//vmc/tmp_dillex.o ../../build//color.o ../../build//textutil.o ../../build//namelist.o ../../build//files.o ../../build //extra.o ../../build//bytestring.o ../../build//common.o ../../build//intlist.o ../../build//dilshare.o -lfl -g -o ../../bin/vmc ../../build//dilshare.o: In function dilprg::dilprg(unit_data, int)': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:84: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:85: undefined reference to dil_list' ../../build//dilshare.o: In function dilprg::~dilprg()': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:132: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:134: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:137: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:138: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:143: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:147: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:148: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:175: undefined reference to dil_free_template(diltemplate, int)'

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-666763017, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXTEYKPULFTZTHLCOWTR6H3XZANCNFSM4PHNIISA .

[Seifert69]

I've begun some very low level memory debugging. In that process ive emplaced the dilprg class. As soon as I did that the server stayed up for 101 days until I stopped it. And zero memory leakage.

Bad news is of course that there's still an unfound mem over run.

I'll keep at it :)

On Sat, Aug 1, 2020, 00:29 Michael Seifert michael@corleone.com wrote:

So I found the memory leak. stack.init(10) apparently was the culprit. I've never seen that class before.

Anyway there seems to be one pesky, elusive bug I'm still hunting for. I'll let you know when I've fixed it, then it should be all good.

Best,

On Fri, Jul 31, 2020 at 8:03 AM Michael Seifert michael@corleone.com wrote:

Blast it, I keep forgetting the Makefile, sorry. It's up now :)

On Fri, Jul 31, 2020 at 1:02 AM Rich Gautier notifications@github.com wrote:

Cannot compile on new pull:

g++ -I. -I.. -I vmc -DVMC -DLINUX -DPOSIX -pipe -m64 -std=c++11 -DMUD_DEBUG -g -Wall -O0 ../../build//vmc/vmc.o ../../build//vmc/vmc_process.o ../../build//v mc/utility.o ../../build//vmc/structs.o ../../build//vmc/destruct.o ../../build/ /vmc/spec_assign.o ../../build//vmc/money.o ../../build//vmc/pp1.o ../../build// vmc/pp2.o ../../build//vmc/pp3.o ../../build//vmc/pp4.o ../../build//vmc/pp5.o . ./../build//vmc/pp6.o ../../build//vmc/pp7.o ../../build//vmc/pp8.o ../../build/ /vmc/vmc_db_file.o ../../build//vmc/tmp_vmcpar.o ../../build//vmc/tmp_vmclex.o . ./../build//vmc/tmp_dilpar.o ../../build//vmc/tmp_dillex.o ../../build//color.o ../../build//textutil.o ../../build//namelist.o ../../build//files.o ../../build //extra.o ../../build//bytestring.o ../../build//common.o ../../build//intlist.o ../../build//dilshare.o -lfl -g -o ../../bin/vmc ../../build//dilshare.o: In function dilprg::dilprg(unit_data, int)': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:79: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:84: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:85: undefined reference to dil_list' ../../build//dilshare.o: In function dilprg::~dilprg()': /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:130: undefined reference to g_nDilPrg' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:132: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:134: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:136: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:137: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:138: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:143: undefined reference to dil_list' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:147: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:148: undefined reference to dil_list_nextdude' /home/rgautier/foxmud/DikuMUD3/vme/src/dilshare.cpp:175: undefined reference to dil_free_template(diltemplate, int)'

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-666763017, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXTEYKPULFTZTHLCOWTR6H3XZANCNFSM4PHNIISA .

[rgautier]

Can you push the new dilprg class? I pulled the one yesterday, but still had an issue. One interesting thing to mention is I'm running in a VPS environment with a 512MB limit. Looking at my history, it looks like I'm hitting memory limits.

Unrelated to your bug, but part of my problem in giving you more data: On the downside, I'm no CentOS expert, but that's the image I'm stuck on, and I STILL can't figure out why I'm not getting core dumps. I've tried removing aprt and still can't get permissions to modify kernel.core_pattern. While my hard limit is 'unlimited' and I've tried setting the soft limit, it constantly resets to 0, and my core_pattern STILL wants to send dumps to the not-installed aprt. Tried installing aprt again to see if I could get it to actually do something...no cores...

So I have to catch it crashing live...with it being unpredictable (12 hours? 36?) I'm not having any luck doing that.

[Seifert69]

Finally found the bug(s). Please pull and try again. I'm quite confident this was it. I ran it super stable for 220 days with no issues and no memory increase.

The GitHUB distro uses 26MB of memory as far as I recall, now that its been fixed. Before it would just keep climbing until it got zapped by your OS.

I have the same core dump gripes as you on Ubuntu and on our production Ubuntu as well

Let me know how it goes. I'd love to hear about your project someday.

Best, Michael

On Sat, Aug 1, 2020 at 6:12 PM Rich Gautier notifications@github.com wrote:

Can you push the new dilprg class? I pulled the one yesterday, but still had an issue. One interesting thing to mention is I'm running in a VPS environment with a 512MB limit. Looking at my history, it looks like I'm hitting memory limits.

Unrelated to your bug, but part of my problem in giving you more data: On the downside, I'm no CentOS expert, but that's the image I'm stuck on, and I STILL can't figure out why I'm not getting core dumps. I've tried removing aprt and still can't get permissions to modify kernel.core_pattern. While my hard limit is 'unlimited' and I've tried setting the soft limit, it constantly resets to 0, and my core_pattern STILL wants to send dumps to the not-installed aprt. Tried installing aprt again to see if I could get it to actually do something...no cores...

So I have to catch it crashing live...with it being unpredictable (12 hours? 36?) I'm not having any luck doing that.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-667553807, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXQQ6NUOINL52DGTPGLR6Q5FVANCNFSM4PHNIISA .

[rgautier]

Pulled and rebuilt now - will leave it up and see how it goes.

For reference tomorrow morning: PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 22553 rgautier 20 0 56160 2108 1512 R 0.0 0.4 0:00.09 top 22518 rgautier 20 0 135192 3348 2800 S 0.0 0.6 0:00.00 ../bin/mplex -w -p 4280 22502 rgautier 20 0 159128 29660 3176 S 0.3 5.7 0:01.02 ../bin/vme

[Seifert69]

Great. Found a few places where I had been too strict, so you may crash it with certain player actions. But it's fixed and you can pull it whenever you want.

On Sun, Aug 2, 2020, 01:43 Rich Gautier notifications@github.com wrote:

Pulled and rebuilt now - will leave it up and see how it goes.

For reference tomorrow morning: PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 22553 rgautier 20 0 56160 2108 1512 R 0.0 0.4 0:00.09 top 22518 rgautier 20 0 135192 3348 2800 S 0.0 0.6 0:00.00 ../bin/mplex -w -p 4280 22502 rgautier 20 0 159128 29660 3176 S 0.3 5.7 0:01.02 ../bin/vme

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#issuecomment-667600988, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXWUIC32V7LEKLTQ3QLR6SSDDANCNFSM4PHNIISA .

[rgautier]

Overnight comparison. Holding steady at 159128 virtual memory overnight. No unusual memory growth.

22974 root 20 0 1029244 41160 29668 S 0.0 7.9 1:23.27 /usr/bin/python2 -s /usr/bin/fail2ban-server -xf start 22518 rgautier 20 0 135192 3364 2816 S 0.0 0.6 0:00.60 ../bin/mplex -w -p 4280 22502 rgautier 20 0 159128 31380 3188 S 0.4 6.0 1:51.44 ../bin/vme

I think you can close the issue 👍

[Seifert69]

I've added a new wiz command "wtimewarp" (w for wiz). You can specify for how many seconds of ticks the game will accelerate.

E.g. wtimewarp 3600 will make vme run at top speed, and what would normally have taken an hour to pass, happens in a few seconds. You can't log in while this is going on. Could be fixed but I didn't bother since it's a debugging function. So now you can simulate the server being up for 36 hours easily rather than having to wait 36 hours.

--Michael

On Sun, Aug 2, 2020 at 2:30 PM Rich Gautier notifications@github.com wrote:

Closed #11 https://github.com/Seifert69/DikuMUD3/issues/11.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Seifert69/DikuMUD3/issues/11#event-3613303147, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOOOMXSWM2UGZHN4UUYHBFTR6VL5ZANCNFSM4PHNIISA .