MAX_DATA hard coded somewhere? <= was: oversized [ports] section
I bumped up the MAXMSG_DATA and MAXMSG_STATUS again, to 768 from 512 (was 256 by default). restarted hobbit ok. the list of SHM via ipcs (appended below) shows the 768K segment is created. However, an hour later, I still ran into messages for hobbitd like this: Oversize data/client msg from 10.77.9.106 truncated (n=722041, limit 524288) This causes all sorts of problems: PROCS/MSGS/FILES got status CLEAR/Non-reporting or false alerts due to required processes are truncated from listings, plus rrddata not gathered for count of PROCS, etc.
Questions
- is MAX_DATA / MAX_STATUS hard-coded at 512K, somewhere?
- any way to get around this ? The super-sized [ports] is here to stay for a few months at least.
---- "hobbitservers.cfg"---cut---- MAXMSG_CLIENT = 1024 MAXMSG_DATA = 768 MAXMSG_STATUS=768
------ Shared Memory Segments -------- key shmid owner perms bytes nattch status 0x0101f2e0 2129920 hobbit 600 786432 2 0x0201f2e0 2162689 hobbit 600 786432 2 0x0301f2e0 2195458 hobbit 600 786432 2 0x0401f2e0 2228227 hobbit 600 262144 2 0x0501f2e0 2260996 hobbit 600 262144 1 0x0601f2e0 2293765 hobbit 600 32768 1 0x0701f2e0 2326534 hobbit 600 524288 2 0x0801f2e0 2359303 hobbit 600 524288 2
For regular user 'hobbit', ulimit don't have any limit set at 512K. [hobbit at saturn ~]$ ulimit -aH core file size (blocks, -c) unlimited data seg size (kbytes, -d) unlimited file size (blocks, -f) unlimited pending signals (-i) 1024 max locked memory (kbytes, -l) 32 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 stack size (kbytes, -s) unlimited cpu time (seconds, -t) unlimited max user processes (-u) 16383 virtual memory (kbytes, -v) unlimited file locks (-x) unlimited [hobbit at saturn ~]$ ulimit -aS core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited file size (blocks, -f) unlimited pending signals (-i) 1024 max locked memory (kbytes, -l) 32 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 stack size (kbytes, -s) 10240 cpu time (seconds, -t) unlimited max user processes (-u) 16383 virtual memory (kbytes, -v) unlimited file locks (-x) unlimited
On Tue, Aug 08, 2006 at 10:56:15AM -0400, Jerry Yu wrote:
I bumped up the MAXMSG_DATA and MAXMSG_STATUS again, to 768 from 512 (was 256 by default). restarted hobbit ok. the list of SHM via ipcs (appended below) shows the 768K segment is created. However, an hour later, I still ran into messages for hobbitd like this: Oversize data/client msg from 10.77.9.106 truncated (n=722041, limit 524288)
------ Shared Memory Segments -------- key shmid owner perms bytes nattch status 0x0101f2e0 2129920 hobbit 600 786432 2 0x0201f2e0 2162689 hobbit 600 786432 2 0x0301f2e0 2195458 hobbit 600 786432 2 0x0401f2e0 2228227 hobbit 600 262144 2 0x0501f2e0 2260996 hobbit 600 262144 1 0x0601f2e0 2293765 hobbit 600 32768 1 0x0701f2e0 2326534 hobbit 600 524288 2 0x0801f2e0 2359303 hobbit 600 524288 2
These ipcs numbers show that your current settings are:
MAXMSG_STATUS=768 MAXMSG_STACHG=768 (matches the _STATUS value) MAXMSG_PAGE=768 (ditto) MAXMSG_DATA=256 MAXMSG_NOTES=256 MAXMSG_ENADIS=32 MAXMSG_CLIENT=512 MAXMSH_CLICHG=512 (matches the _CLIENT setting)
---- "hobbitservers.cfg"---cut---- MAXMSG_CLIENT = 1024 MAXMSG_DATA = 768 MAXMSG_STATUS=768
Hmm - try removing that space around the '=' sign:
MAXMSG_CLIENT=1024 MAXMSG_DATA=768 MAXMSG_STATUS=768
I think that's the problem.
Regards, Henrik
removed the space in hobbitserver.cfg. now it looks better. Thanks again, Henrik.
------ Shared Memory Segments -------- key shmid owner perms bytes nattch status 0x0101f2e0 2621440 hobbit 600 786432 2 0x0201f2e0 2654209 hobbit 600 786432 2 0x0301f2e0 2686978 hobbit 600 786432 2 0x0401f2e0 2719747 hobbit 600 786432 2 0x0501f2e0 2752516 hobbit 600 262144 1 0x0601f2e0 2785285 hobbit 600 32768 1 0x0701f2e0 2818054 hobbit 600 1048576 2 0x0801f2e0 2850823 hobbit 600 1048576 2
One tiny problem. had to manually remove shm/sem. w/o doing so, restarting hobbitd failed with errors similar to bumping into 'ulimit'. It is 4.2-RC-20060712 w/o patch on a CentOS 4.3 server. 2006-08-08 12:21:51 Could not get shm of size 1048576: Invalid argument 2006-08-08 12:21:51 Cannot setup clichg channel 2006-08-08 12:21:51 Task hobbitd terminated, status 1
On 8/8/06, Henrik Stoerner <henrik at hswn.dk> wrote:
I bumped up the MAXMSG_DATA and MAXMSG_STATUS again, to 768 from 512 (was 256 by default). restarted hobbit ok. the list of SHM via ipcs (appended below) shows the 768K segment is created. However, an hour later, I still ran into messages for hobbitd like this: Oversize data/client msg from 10.77.9.106 truncated (n=722041, limit
On Tue, Aug 08, 2006 at 10:56:15AM -0400, Jerry Yu wrote: 524288)
------ Shared Memory Segments -------- key shmid owner perms bytes nattch status 0x0101f2e0 2129920 hobbit 600 786432 2 0x0201f2e0 2162689 hobbit 600 786432 2 0x0301f2e0 2195458 hobbit 600 786432 2 0x0401f2e0 2228227 hobbit 600 262144 2 0x0501f2e0 2260996 hobbit 600 262144 1 0x0601f2e0 2293765 hobbit 600 32768 1 0x0701f2e0 2326534 hobbit 600 524288 2 0x0801f2e0 2359303 hobbit 600 524288 2
These ipcs numbers show that your current settings are:
MAXMSG_STATUS=768 MAXMSG_STACHG=768 (matches the _STATUS value) MAXMSG_PAGE=768 (ditto) MAXMSG_DATA=256 MAXMSG_NOTES=256 MAXMSG_ENADIS=32 MAXMSG_CLIENT=512 MAXMSH_CLICHG=512 (matches the _CLIENT setting)
---- "hobbitservers.cfg"---cut---- MAXMSG_CLIENT = 1024 MAXMSG_DATA = 768 MAXMSG_STATUS=768
Hmm - try removing that space around the '=' sign:
MAXMSG_CLIENT=1024 MAXMSG_DATA=768 MAXMSG_STATUS=768
I think that's the problem.
Regards, Henrik
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
On Tue, Aug 08, 2006 at 12:34:46PM -0400, Jerry Yu wrote:
removed the space in hobbitserver.cfg. now it looks better. Thanks again,
One tiny problem. had to manually remove shm/sem. w/o doing so, restarting hobbitd failed with errors similar to bumping into 'ulimit'. It is 4.2-RC-20060712 w/o patch on a CentOS 4.3 server. 2006-08-08 12:21:51 Could not get shm of size 1048576: Invalid argument 2006-08-08 12:21:51 Cannot setup clichg channel
The original RC version failed to clean up this shmem segment when shutting down, causing this problem the next time you started Hobbit. It was fixed some time ago in the beta-patches.
Regards, Henrik
suspected so. built new RPMs usign all-in-one patch dated 08/07. The default upgrade behavior seems to be more disrupting than I expect. I guess I'll need to modify the RPM spec file to make it a bit friendly to upgrade, plus to make it SELINUX aware.
On 8/8/06, Henrik Stoerner <henrik at hswn.dk> wrote:
On Tue, Aug 08, 2006 at 12:34:46PM -0400, Jerry Yu wrote:
removed the space in hobbitserver.cfg. now it looks better. Thanks again,
One tiny problem. had to manually remove shm/sem. w/o doing so, restarting hobbitd failed with errors similar to bumping into 'ulimit'. It is 4.2-RC-20060712 w/o patch on a CentOS 4.3 server. 2006-08-08 12:21:51 Could not get shm of size 1048576: Invalid argument 2006-08-08 12:21:51 Cannot setup clichg channel
The original RC version failed to clean up this shmem segment when shutting down, causing this problem the next time you started Hobbit. It was fixed some time ago in the beta-patches.
Regards, Henrik
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
participants (2)
-
henrik＠hswn.dk
-
jjj863＠gmail.com