opcragt not working, but opcdeploy works fine

[Bearwhale]

Well, another perplexing problem has come up, guys.

I'm trying to get the status of the server. When I type:

root_server:/> opcragt -status node_name

I get:

"Node node_name:

Cannot get status information from node node_name. (OpC40-428)
Network communication problems occurred. (OpC40-427)

-------------------------------------------------------------------------------
CTRL - CommunicationException:
-------------------------------------------------------------------------------
(ctrl-21) Communication error when executing 'Status' method.
(xpl-316) connect() to '[IP address of the server]' failed.
(RTL-238) Connection timed out (OpC40-2130)
The connection timed out because either the node is down or there's
a problem with the network and the node can't be reached. (OpC40-2171)
Failed."

Note: I edited out the node name, the root server name and the IP address.

What's odd, though, is that I can use the "opcdeploy" command to get the status of the node.

When I type: 

root_server:/> opcdeploy -cmd "ovc -status" -node node_name

I get:

"ovcd OV Control CORE (1184) Running
opcle OVO Logfile Encapsulator AGENT,EA (6700) Running
coda OV Performance Core COREXT (6900) Running
opcmona OVO Monitor Agent AGENT,EA (6996) Running
opcacta OVO Action Agent AGENT,EA (7284) Running
opcmsga OVO Message Agent AGENT,EA (6728) Running
opcmsgi OVO Message Interceptor AGENT,EA (8156) Running
ovbbccb OV Communication Broker CORE (5656) Running
ovconfd OV Config and Deploy COREXT (5764) Running"

I'm confused... why can't I get the information from the remote agent but I CAN get it from issuing a command to the node?

If you guys could help me figure this out, that'd be awesome!

[Gino Castoldi_2]

Hi,

I have seen this type of error before. I would recommend adding these managed node configuration variables to one server as a test to see if it helps or not.

1) First  stop the server's om agent.

2) Add in these agent configuration variables.

/opt/OV/bin/ovconfchg  -ns  bbc.cb  -set  SERVER_BIND_ADDR      <node ip>

/opt/OV/bin/ovconfchg  -ns  bbc.http  -set  SERVER_BIND_ADDR    <node ip>

/opt/OV/bin/ovconfchg  -ns  bbc.http  -set  CLIENT_BIND_ADDR       <node ip>

/opt/OV/bin/ovconfchg   -ns  eaagt  -set   OPC_IP_ADDRESS             <node ip>

/opt/OV/bin/ovconfchg   -ns  eaagt  -set   RESOLVE_ADDRESS_IP   <mgmt server ip>

3) Restart the om agent.

4) Check for any errors (System.txt logfile, etc.).

5) Run opcragt -status <mgd_node>  again and check the results.

I believe the reason for this is that the agent gets confused as to what its primary ip address is because in the local agent configuration file it is listed as "localhost".

-Gino

[Bearwhale]

Hmm... I tried your suggestion but it doesn't seem to have fixed the problem. Anything else I could do?

[Bearwhale · Accepted solution]

Weird. I at least fixed the problem by following these steps (heavy-handed though they may be):

1. Remove agent from server.
2. Make sure no traces of agent remain.
3. Delete agent from OVO.
4. Re-install agent on server.
5. Open motif and "Add Node to OVO" in Certificate Requests.

6. Run custom script to read and issue certificates.

Not sure if this will help anyone else, but that's how I fixed the problem.

[KBerger]

Is this a standalone OM server or does in run in a cluster?