"netstat -in" or "netstat -rn" will hang

[jack wu_2]

ES45 V5.1A cluster PK4
I run "netstat -in" or "netstat -rn"
the command will hang, and no output.
mes1pdb2# dbx -k /vmunix
dbx version 5.1
Type 'help' for help.

thread 0xfffffc017fe12a80 stopped at [thread_block:3230 ,0xfffffc00004eaf00] Source not available

warning: Files compiled -g3: parameter values probably wrong
(dbx) p msginfo
struct {
msgmax = 8192
msgmnb = 16384
msgmni = 512
msgtql = 40
msg = 1635169
}
(dbx) p cur_msghdrs
40
(dbx) quit

I want ot know the reason of msg-queue will be full.

thanks

[Joris Denayer]

jack,

In your entry, I remark that you are aware of a possible netstat problem when
msgtql = cur_msghdrs

You can use the command ipcs -qa
This will show the PID of the processes holding these messagequeue's

A kill of these processes could help.
But, I think you should first investigate why this situation occurs.

I suggest also to increase the value of msgtql in /etc/sysconfigtab (in the ipc-subsystem)

A reboot will be required, as this is no dynamic parameter. %-{

Rgrds

Joris

[jack wu_2]

Hi joris
I can't confirm PID of messages queue.
LSPID? LRPID?

use dbx to trace netstat, but no more info to analyze why the cur_msghdrs will be full.

the log file is attached.

thanks
jack

[Ralf Puchner]

jack,

is there also a problem calling arp?

[Joris Denayer]

Hi ,

According the manpages.
LSPID (a,p)
The ID of the last process that sent a message to the associated queue.

LRPID (a,p)
The ID of the last process that received a message from the associated
queue.

QNUM (a,o)
The number of messages currently outstanding on the associated message
queue.


The second line of the messagequeues gives
q 1537 0x4c544952 --rw-rw-rw- root daemon root daemon 8320 40 16384 1504259 1341906 21:25:08 21:25:08 17:00:51

40 is the number of currently outstanding messages in the messagequeue with ID 1537

1504259 is the PID of the process that sent the last message to the messagequeue

1341906 is the PID of the last process that received a message


Joris