Originally posted by: SystemAdmin

AIX 5.3
oslevel -r responds: 5300-05
Model: 9131-52A

Has been running fine for years. Just recently, we have begun generating multiple "respawn too rapid" errors in the error log as below:

---

LABEL: INIT_RAPID
IDENTIFIER: 3A30359F

Date/Time: Wed Jul 6 06:27:16 PDT 2011
Sequence Number: 11511
Machine Id: 00047500D700
Node Id: nwwinc
Class: S
Type: TEMP
Resource Name: init

Description
SOFTWARE PROGRAM ERROR

Probable Causes
SOFTWARE PROGRAM

User Causes
PERFORMANCE DEGRADED

Recommended Actions
REVIEW DETAILED DATA

Detail Data
SOFTWARE ERROR CODE
Command is respawning too rapidly. Check for possible errors.
COMMAND
id: agd "/usr/appgen/agdaemon.sh > /home/sysadmin/agd.out 2>&1"

---

The only changes to the system or environment are switching all the PCs to Win7.

This program has been running reliably on this and previous levels of AIX plus this and previous RISC systems for some time. It provides a client login to an accounting application on AIX 5.3 from an ethernet local WAN made of Cisco 831-K9 routers. There are no serial terminals to create this problem from a TTYHOG overrun. There are no recent AIX updates. There is no significant error output sent to the error output location.

There seems to be no AIX troubleshooting information that really explains what causes "respawning too rapidly" errors.

Any ideas or referrals to decent technical documentation for "respawn too rapid" errors would be appreciated.
#AIX-Forum

Originally posted by: shargus

"respawning too rapidly" basically means exactly that - the process dies, init respawns it, repeat... When this happens too many times in certain period of time (5 minutes, I think), init quits trying and issues the message.

Try running the command from a root shell. I'll bet you'll get a more informative message.
You may have to disable it in inittab first.
#AIX-Forum

Originally posted by: SystemAdmin

THX for the response. Starting from the command line may reveal something, so I appreciate the suggestion.

What I am trying to find out is; can I "nice" it to a lower priority or something to reduce the rapid respawn. It has run fine for years. This is not the only command that I have seen become a respawn problem on this and other systems. As such, there should be some priority lowering or respawn delay that would cut it down. Maybe I should embed a "sleep 2" in the shell that invokes it. Do you have any idea if that might help??
#AIX-Forum

Originally posted by: shargus

According to the man page for init,

If the init command finds that it is continuously running an entry in the /etc/inittab file (more than five times in 225 seconds), it assumes that an error in the entry command string exists. It then prints an error message to the console and logs an error in the system error log. After the message is sent, the entry does not run 

for 60 seconds. If the error continues to occur, the command will respawn the entry only five times every 240 seconds. The init command continues to assume an error occurred until the command does not respond five times in the interval, or until it receives a signal from a user. The init command logs an error 

for only the first occurrence of the error.

So, there is no way to change the behavior of init. Adding a "sleep 2" won't change anything, either (will still respawn 5 times in 225 seconds). You COULD sleep for a really long time (4 minutes should do the trick), but it still doesn't fix the underlying problem - your process starts, dies, then four minutes later it restarts, then dies again...

Renicing the process - doesn't really matter, unless it's eating up a significant amount of CPU time.

You didn't mention it in your first post, but have you checked the agd.out log file?

I still think your best bet is to start the process up from a root shell, then see what happens when it dies.
#AIX-Forum

Originally posted by: SystemAdmin

THX, again, for the response.

The results for agd.out are typically this:

1. cat agd.out

server: can't bind local address
Using default port (8188)

The process being respawned is bound to port 8188 to perform login to the accounting application. Other users are likely
logged on at the time of the respawning and already utilizing a port 8188 connection. I have no way to have the script
clear the port (8188) prior to the attempt to bind it.

This has become an issue since all the PCs on the network (and they are the devices logging in via this port) converted to
Win7. I prefer not to tell a customer "Oh well, it's Windows' fault...live with it". On the other hand, the AIX servers have
run reliably for over 10 years running this client login process. Though diags show nothing, I can only guess an intermittent
ethernet problem if it's the AIX box.
#AIX-Forum