[SNMP4J] 1 sec to send a pdu with SnmpRequest

PHIL BERGSTRESSER phil.bergstresser at adtran.com
Fri Jul 27 15:14:47 CEST 2007


Jason,
    I would predict that the delay is in the agent responsiveness. The
way you evaluate that is to put a sniffer on the wire that reports
message traffic and timestamps and then you will have clear evidence of
the responsiveness of the agent. 
    If you are running your test routines on a Solaris based system,
there is a built in program named snoop that will trace the ethernet
interface out and into your machine with timestamps, and protocol
interpretation along with the hex packet dump. If on MS Windows,
download the free Ethereal packet sniffer. This presumes you are using
Ethernet as your interface. 
    UDP or TCP packets can be isolated by protocol so you don't have to
look at other network traffic, but you want to run wide open at first to
see if there are other messages that are consuming bandwidth that might
be interfering.
    HTH,
Phil

> -----Original Message-----
> From: snmp4j-bounces at agentpp.org 
> [mailto:snmp4j-bounces at agentpp.org] On Behalf Of Jason Grant
> Sent: Friday, July 27, 2007 2:03 AM
> To: snmp4j at agentpp.org
> Subject: [SNMP4J] 1 sec to send a pdu with SnmpRequest
> 
> I've written my test routines to confirm proper operation of 
> my agent by sending PDUs using 
> org.snmp4j.tools.console.SnmpRequest.  Unfortunately, the 
> send method of this class is taking over 1 second to transmit 
> and receive a response, making table traversal very slow.  I 
> have not profiled this class, nor the agent, which is based 
> on TestAgent. 
> 
> I'm a little concerned that the bottleneck may be in the 
> agent itself, rather than the client.  Tips about how to 
> reduce the latency would be appreciated, since if the 
> sluggishness is in the agent, it may be a little slow to use 
> in our production environment.
> 
> Thanks,
> 
> Jason.
> 
> -----------------
> Utiba Pty Ltd
> This message has been scanned for viruses and dangerous 
> content by Utiba mail server and is believed to be clean.
> 
> _______________________________________________
> SNMP4J mailing list
> SNMP4J at agentpp.org
> http://lists.agentpp.org/mailman/listinfo/snmp4j
> 



More information about the SNMP4J mailing list