Hi Daphna,
First of all, creating a dedicated thread for each
target is not a good idea. It takes up unnecessary
resources and slows down the application. More worse,
it can create other problems, for example, if
you do not share transport mappings.
To send out messages, you can use a single transport
mapping per protocol. If you create and close (hopefully)
transport mappings often, you can get many ports on
your machine in CLOSE_WAIT state and run out of ports.
Your particular problem could be caused by something
similar. If the client is shutdown ungracefully,
then the TCP connection hold by the notification
originator is half-closed and the server will probably
not recognize it until it send a packet that times
out after the TCP default timeout.
This issues is networking related and not related to
SNMP4J.
Best regards,
Frank
Daphna Wasserman (dwasserm) wrote:
Hi,
Our application forwards SNMP V2 traps to different clients listening on
either TCP or UDP. For each combination of address+port we create a
dedicated thread, and if the connection is of type TCP, we create an
instance of DefaultTCPTransportMapping and use it to send messages to
the specific client.
Currently we are using snmp4j.jar version 1.7.1 + java 1.5.0_08-b03
We have an issue in which after we restart the client (while the server
keeps trying to send traps to the client) , the new client instance does
not receive traps from the server any longer. We see the following
message at the server log:
May 04 2008 16:10:02.469 IDT Snmp4JLogAdapter.debug DEBUG
{64.103.124.103/50004=SocketEntry[peerAddress=64.103.124.103/50004,socke
t=Socket[unconnected],lastUse=Sun May 04 14:56:08 IDT 2008]}
May 04 2008 16:10:03.339 IDT Snmp4JLogAdapter.debug DEBUG Looking up
connection for destination '64.103.124.103/50004' returned:
SocketEntry[peerAddress=64.103.124.103/50004,socket=Socket[unconnected],
lastUse=Sun May 04 14:56:08 IDT 2008]
We've tried upgrading snmp4j to 1.8.2 and even to 1.9.1e with no
improvement - after the client was restarted traps were still not
received, and the following error appeared in the server log:
May 04 2008 11:46:35.015 IDT Throwable.printStackTrace ERROR
SOut: java.lang.Exception: Stack trace
May 04 2008 11:46:35.015 IDT Throwable.printStackTrace ERROR
SOut: at java.lang.Thread.dumpStack(Thread.java:1158)
May 04 2008 11:46:35.015 IDT Throwable.printStackTrace ERROR
SOut: at
com.cisco.nm.cmp.net.protocols.snmp.Snmp4JLogAdapter.dumpFurtherInfo(Snm
p4JLogAdapter.java:152)
May 04 2008 11:46:35.015 IDT Throwable.printStackTrace ERROR
SOut: at
com.cisco.nm.cmp.net.protocols.snmp.Snmp4JLogAdapter.error(Snmp4JLogAdap
ter.java:57)
May 04 2008 11:46:35.015 IDT Throwable.printStackTrace ERROR
SOut: at
org.snmp4j.transport.DefaultTcpTransportMapping$ServerThread.processPend
ing(Unknown Source)
May 04 2008 11:46:35.015 IDT Throwable.printStackTrace ERROR
SOut: at
org.snmp4j.transport.DefaultTcpTransportMapping$ServerThread.run(Unknown
Source)
May 04 2008 11:46:35.017 IDT Throwable.printStackTrace ERROR
SOut: at java.lang.Thread.run(Thread.java:595)
May 04 2008 11:46:35.017 IDT Throwable.printStackTrace ERROR
SOut: at
org.snmp4j.util.DefaultThreadFactory$WorkerThread.run(Unknown Source)
May 04 2008 11:46:35.017 IDT Snmp4JLogAdapter.debug DEBUG Firing
transport state event:
org.snmp4j.transport.TransportStateEvent[source=org.snmp4j.transport.Def
[EMAIL PROTECTED],peerAddress=64.103.124.103/50004,newSta
te=4,cancelled=false,causingException=java.nio.channels.ClosedChannelExc
eption]
May 04 2008 11:46:35.017 IDT Snmp4JLogAdapter.error ERROR
java.nio.channels.ClosedChannelException
We'll appreciate your help
Thanks
Daphna
_______________________________________________
SNMP4J mailing list
[email protected]
http://lists.agentpp.org/mailman/listinfo/snmp4j
--
AGENT++
http://www.agentpp.com
http://www.mibexplorer.com
http://www.mibdesigner.com
_______________________________________________
SNMP4J mailing list
[email protected]
http://lists.agentpp.org/mailman/listinfo/snmp4j
