Seeing errors "HPUX Error: 233: No buffer space available"

[Timothy Czarnik]

Hey all,

We're seeing messages like this in our Oracle listener.log. When I do some research on the error it points to a memory resource issue on the server but from what I can see the server is not having a memory resource issue. Has anyone seen this issue before?

Tim

19-MAR-2007 12:27:24 * 12560
TNS-12560: TNS:protocol adapter error
TNS-00530: Protocol adapter error
HPUX Error: 233: No buffer space available

[Julio Yamawaki]

Hi,

Read this message from metalink:

fact: Oracle Net
fact: TNS Listener
fact: HP 9000 Series HP-UX
symptom: Connections to database hang
symptom: Errors reported in the listener.log
symptom: TNS-12560: TNS:protocol adapter error
symptom: TNS-00530: Protocol adapter error
symptom: HPUX Error: 233: No buffer space available
cause: Problem is generally seen during high workload on machine.
The underlying problem of these errors is the operating system.
The HP UX system file /usr/include/sys/errno.h shows the following information
for the OS error 233:

# define ENOBUFS 233 /* No buffer space available */

ENOBUFS means there is no sufficient memory available and the system(kernel)
can not allocate any more.
Application will usually retry the operation when it detects this error from a
system call since it indicates there is a transient resource shortage.
It is the Operating system that refuses the resource request from the listener.
The virtual memory allocation routine of the OS will determine if a swap can be
made to disk of a real memory segment thereby allowing the listener access to
some more real memory

fix:

Free up memory on the machine

Regards,

[Timothy Czarnik]

Thanks for the quick response. I've seen that metalink article and it points to a system memory issue. The system is not memory bound so I'm wondering if anyone has seen this issue before.

[Steven E. Protter]

Shalom,

You can cause this issue by allocating too much space to the system buffer cache.

The default is dbc_max_pct=50 and dbc_min_pct=5.

I generally lower this in 11.11 systems to max set under 10. This frees up a lot of memory which the Oracle SGA can then use.

SEP

[rick jones]

99 times out of 10 and ENOBUFS on an accept() call means that by the time the server application got around to calling accept() to accept the connection, the remote client gave-up and closed/aborted it. IE it means the server application was too slow relative to the client's expectations.

One has to either reset the client application's expectations, or tune and improve the performance of the server application.