Console Error Message

[Kashif Naseer_1]

My server is running fine, everything is just fine but my server is displaying an error message on console again and again after some time
"ln: failed to get new buffer for received packets"

Server is rp7420 connected in SAN with EVA5000.

[Peter Godron]

Kashif,
is this a system error message or an application error message?

[Cem Tugrul]

need more info...

[RAC_1]

Is this the only message?? Is there additional text with it??
Check syslog.log
dmesg

[Sยภเl Kย๓คг]

Hi,

From when are u getting this error maessage, what is the last change made to the system, both hardware and software.

Check syslog for any possible entries related to this error.

Regards,
Sunil

[Sยภเl Kย๓คг]

Hi,

Mostly this is something related to network

Regards,
Sunil

[Kashif Naseer_1]

I am able to login the system and there is no message in syslog.log and dmesg.
Server is sending message on console after every 10-15 seconds. Oracle & SAP is running on the system.

Regards,

Kashif

[Steven E. Protter]

Shalom Kashif,

Seems that you do have networking issues.

It would be a good idea to log onto the console and do a quick check of network functionality.

This message means that things are not running fine.

It could be a problem outside your machine, or it could be the prelude to a network card failure.

It would be a good idea to run some hardware diagnostics on networking with cstm,mstm or xstm, your choice, whichever tool you prefer.

SEP

[Sยภเl Kย๓คг]

Hi,

Did u notice any problem related to the network performance by the system.

Check the status of the NIC LEDs behind teh server, have close observation for at least 30Mins.

How is your network and backbone structure.

regards,
Sunil

[Sameer_Nirmal]

Hi,

The "buffer" referred in the message could be TCP buffer which couldn't be allocated for the received packets.

I doubt as why the error is being given by "ln". Are you using NFS? If yes, which
version?

I would check the patch status of the system
for nfs and network related patches first.

I would go to debug network activity using
# netstat -i
# netstat -s
# netstat -m

Enabling network tracing for a short time using "nettl" and them check the logs using
"netfmt"

Checking network parameter using ndd and looking for possiblity tune parameters related to tcp buffer.