[SMEG] ZhayTee Posted April 11, 2002 Share Posted April 11, 2002 We're running the new 1.02b server in Linux, and we're still getting random crashes. We have no idea what the crashes are based upon, though the logs may shed some light on the subject. I've been able to save two logs from seperate crashes, and at the point of the crash, the lines are similar. No idea if this is a meaningless coincidence, but. . . check it out: jk2.crash.log ------------ 834:16 Kill: 14 10 3: m0ggY killed Camel by MOD_SABER 834:18 Item: 14 item_medpak_instant 834:19 Item: 7 weapon_bowcaster 834:20 Item: 19 item_shield_sm_instant 834:20 Item: jk2.crash2.log ------------- 758:12 Item: 5 item_medpak_instant 758:13 Item: 5 item_medpak_instant 758:16 Item: 12 item_shield_sm_instant 758:16 Item: 5 item_shield_sm_instant 758:17 So, we're not sure. For all we know, it could be a DoS explot or whatnot. --JT Link to comment Share on other sites More sharing options...
MatrixCPA Posted April 11, 2002 Share Posted April 11, 2002 If the server is running into a segment fault--which is what mine does every so often--it would not make it into the log. To see it, you'd need to be looking at the server console when it happened or have the server running inside screen and not be set to auto-restart. The crash vulnerability produces a segment fault when it occurs. However, a few of us still have a theory about the master server being on the blink sometimes and causing a problem. I might have more data on that later. Link to comment Share on other sites More sharing options...
xmscratch Posted April 12, 2002 Share Posted April 12, 2002 Same problem here. This release is infinitely less stable than it's predecessor, and uses close to twice the system resources. There is absolutely no error during the crash. All processes just plain die. I'm happy to supply specs and details if anyone is so curious. xm-scratch http://www.fragfarm.net scratch@fragfarm.net Link to comment Share on other sites More sharing options...
MatrixCPA Posted April 12, 2002 Share Posted April 12, 2002 I'm going to be testing out a new auto-restart script tonight that will log the seg fault error and ping the master server and log that result. If it works as I expect, I'll post it here so other admins can use it to collect more data about the crashes. Link to comment Share on other sites More sharing options...
xmscratch Posted April 14, 2002 Share Posted April 14, 2002 I finally had a visible error on the last crash: Kill: 1 1 35: [DD]Kimera killed [DD]Kimera by MOD_SUICIDE setteam: 1 FREE SPECTATOR ClientUserinfoChanged: 1 n\[DD]Kimera\t\3\model\jeditrainer/default\g_redteam\Em pire\g_blueteam\Rebellion\c1\4\c2\5\hc\100\w\0\l\0\tt\0\tl\0 ClientBegin: 1 ******************** ERROR: opStack corrupted in compiled code ******************** ----- Server Shutdown ----- Sending heartbeat to masterjk2.ravensoft.com Sending heartbeat to masterjk2.ravensoft.com ==== ShutdownGame ==== ShutdownGame: ------------------------------------------------------------ AAS shutdown. --------------------------- DROPPED _____________ Link to comment Share on other sites More sharing options...
MatrixCPA Posted April 14, 2002 Share Posted April 14, 2002 opStack crashes are frequently caused by using g_dismember "1". Make sure you have that set to "0" instead. Just one more work around. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.