Hello, I am running several SuSE 2.4.7 instances under zVM V4.2. On one of my DB2/UDB V7.2 instances, I have an intermittent (1-2 times weekly) problem where it goes into a loop. This problem just began when we converted from the 2.2.16 to the 2.4.7 Kernel. I can ping the system, but that is about it; telnet sessions and all DB2 connections hang. I logged on to the user in zVM and enter a few quick "#CP D PSW ALL" commands. The PSW always displays either address 800D7120 or 80D7122, so it appears to be a very tight loop.
I used the contents of the memory map, "/boot/System.map-2.4.7-timer-SMP", to determine that this address is in routine "shm_open". This appears to have to do with the "Shared Memory Filesystem". Has anyone seen this? Any ideas would be much appreciated. Thanks, Hank Calzaretta Wallace Computer Services, Inc. ====>Output of "#CP D PSW ALL" CP D PSW ALL PSW = 07081000 800D7120 EXT 1004 18 OLD 07082000 800D7122 58 NEW 04080000 80013212 SVC 0002 20 OLD 070DD000 C0AE53AA 60 NEW 04080000 800128BA PRG 0004 28 OLD 070DC000 C0AE41A2 68 NEW 04080000 80012F08 MCH 0000 30 OLD 00000000 00000000 70 NEW 04080000 8001330A I/O 0087 38 OLD 07080000 80139366 78 NEW 04080000 80013086 ====>Output of "CP D TX D7000-D7130" follows: V00000000 00080000 8001342C 00000000 00000000 06 *......4,........* V000D7000 F0782020 8950000F 18291A2C 1A5B5830 06 *.x .P...).,.[X0* V000D7010 D0180DE1 1AC25820 D020D703 20002000 *......X . .. . .* V000D7020 07F01AAC 19A8A7B4 0006A7C8 000050A0 *..............P.* V000D7030 F0881818 1A1719A1 A7240010 5820D010 *.........$..X ..* V000D7040 A7BA0001 59B02008 A7C4FF8D 5810F084 *....Y. .....X...* V000D7050 A7280001 50201000 5840D008 A7180001 *.(..P [EMAIL PROTECTED] V000D7060 58304000 18231A21 BA324000 A744FFFC [EMAIL PROTECTED]@..D..* V000D7070 1222A734 00065810 D01C1824 0DE11818 *.".4..X....$....* V000D7080 5B10F088 1A911BC1 5810F080 19C75090 *[.......X.....P.* V000D7090 1000A7D4 000318C7 12CCA7B4 0004A7C8 *................* V000D70A0 0000182C 5840F0C8 986FF0A8 07F40707 *...,[EMAIL PROTECTED] V000D70B0 90CFF030 A7D5000C 002D92B0 001D9F84 *...0.....-......* V000D70C0 001DA066 000203C8 0021AB70 181FA7FA *...f.....!.p....* V000D70D0 FFA05010 F0005810 20345850 D0005820 *..P...X. 4XP..X * V000D70E0 10085830 20085820 30281812 1222A7A4 *..X0 .X 0(..."..* V000D70F0 0004A71A 7FFF8A10 000F8910 000F1832 *...............2* V000D7100 1B315930 5000A754 0006A7C8 0000A7F4 *.1Y0P..T........* V000D7110 001C1845 A74A0024 A7150004 83000044 *...E.J.$.......D* V000D7120 1F00BA01 4000A744 FFFB8930 00025820 [EMAIL PROTECTED] * V000D7130 50285813 20001211 A7740006 D7034000 *P(X. [EMAIL PROTECTED] ====>Output of "uname -a" command follows: Linux wcs-mf-winxs-db2p 2.4.7-timer-SMP #1 SMP Thu Apr 25 15:57:41 GMT 2002 s390 unknown ====>Contents of System.map-2.4.7-timer-SMP around address 000D7120: 000d6660 t alloc_undo 000d673c T sys_semop 000d6c2c T sem_exit 000d6ed0 t sysvipc_sem_read_proc 000d70b0 t shm_open 000d71a0 t shm_destroy 000d7220 t shm_close 000d738c t shm_mmap 000d749c t newseg -----Original Message----- From: Gustavson, John (IDS ECCS) [mailto:[EMAIL PROTECTED] Sent: Tuesday, April 01, 2003 9:52 AM To: [EMAIL PROTECTED] Subject: Re: Things not starting at IPL Exactly. S99donothing will be at the end as long as you don't have anything else is S99, after S99d (eg. S99e ~ S99z.) Regards John Gustavson Enterprise Central Software Services (ECSS) 570 Washington Street - 2nd floor New York, New York, 10080-6802 Telephone: 1-212-647-3793 Fax: 1-212-647-3321 Email: [EMAIL PROTECTED] -----Original Message----- From: James Melin [mailto:[EMAIL PROTECTED] Sent: Tuesday, April 01, 2003 10:45 AM To: [EMAIL PROTECTED] Subject: Re: Things not starting at IPL So by 'this runs at the end of all our start up scripts' you mean anything in rc.x that starts with 'S' or somethign else? |---------+----------------------------> | | "Gustavson, John | | | (IDS ECCS)" | | | <[EMAIL PROTECTED]| | | nge.ml.com> | | | Sent by: Linux on| | | 390 Port | | | <[EMAIL PROTECTED]| | | IST.EDU> | | | | | | | | | 04/01/2003 09:36 | | | AM | | | Please respond to| | | Linux on 390 Port| | | | |---------+----------------------------> >--------------------------------------------------------------------------- ---------------------------------------------------| | | | To: [EMAIL PROTECTED] | | cc: | | Subject: Re: Things not starting at IPL | >--------------------------------------------------------------------------- ---------------------------------------------------| We created a script S99donothing in /etc/init.d/rc3.d as follows: case "$1" in start) echo -n "Doing nothing" sleep 3 rc_status -v ;; This runs at the end of all our start up scripts. Regards John Gustavson Enterprise Central Software Services (ECSS) 570 Washington Street - 2nd floor New York, New York, 10080-6802 Telephone: 1-212-647-3793 Fax: 1-212-647-3321 Email: [EMAIL PROTECTED] -----Original Message----- From: James Melin [mailto:[EMAIL PROTECTED] Sent: Tuesday, April 01, 2003 10:30 AM To: [EMAIL PROTECTED] Subject: Re: Things not starting at IPL Could you give some details as to which startup scripts and what specific commands you did to allow this transition ? |---------+----------------------------> | | "Gustavson, John | | | (IDS ECCS)" | | | <[EMAIL PROTECTED]| | | nge.ml.com> | | | Sent by: Linux on| | | 390 Port | | | <[EMAIL PROTECTED]| | | IST.EDU> | | | | | | | | | 04/01/2003 09:17 | | | AM | | | Please respond to| | | Linux on 390 Port| | | | |---------+----------------------------> > ---------------------------------------------------------------------------- --------------------------------------------------| | | | To: [EMAIL PROTECTED] | | cc: | | Subject: Re: Things not starting at IPL | > ---------------------------------------------------------------------------- --------------------------------------------------| We had the same problem with various things not starting intermittently. The processes are getting hup'ed if the boot process hasn't ended and closed the log. Because these processes are writing to the boot log, they get hup'ed too. We put in a dummy script to sleep at the end of the start-up scripts to allow boot to close the log and terminate. This solved the intermittent problem. Actually it was Ken Hall who discovered this. Regards John Gustavson Enterprise Central Software Services (ECSS) 570 Washington Street - 2nd floor New York, New York, 10080-6802 Telephone: 1-212-647-3793 Fax: 1-212-647-3321 Email: [EMAIL PROTECTED] -----Original Message----- From: James Melin [mailto:[EMAIL PROTECTED] Sent: Tuesday, April 01, 2003 9:39 AM To: [EMAIL PROTECTED] Subject: Things not starting at IPL I have a situation where various things are not starting properly at boot. I see the message for the httpd server on the HMC console: Starting httpd [ LDAP PERL PHP4 SSL ] but when I telnet in, the task is not running and I don't see anything in the log that truly indicates why it's not there. Also, JBOSS is not starting at IPL, even though the items in etc/init.d/rc3/d are ok. (or appear to be). Lastly, one component of DB2connect EE isn't starting either. That is the db2sysc process, in the 8.1 version of DB2 connect. I have had to logon as db2inst1 and issue a db2start command to get that to work. Nothing for it was created in /etc/init.d or any of the rc.x sub dirs. Any insight? Here are the RC3.d entries for apache/jboss. I believe the links are correct. The actual scripts themselves do function to start/stop the tasks. Apache lrwxrwxrwx 1 root root 9 Feb 27 13:57 K02apache -> ../apache lrwxrwxrwx 1 root root 9 Mar 9 14:46 S23apache -> ../apache Jboss: lrwxrwxrwx 1 root root 8 Mar 11 13:56 K01jboss -> ../jboss lrwxrwxrwx 1 root root 8 Mar 11 13:56 S21jboss -> ../jboss
