cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1379
Views
0
Helpful
10
Replies

SNASW-3-EVENT logging messages

whecox
Level 1
Level 1

I just configured SNASW to connect a PU to EE on the M/F. If the SWNET is not active on the M/F, the cisco console log fills up with event messages. Is there anyway to filter or disable these SNASW activation attempt logging messages. Every 30 seconds.

SNASW-3-EVENT: Pipe to DLUS activated

SNASW-3-DLUR_LOG_2: PROBLEM

SNASW-3-EVENT: APPN to DLUS deactivated

10 Replies 10

jishi
Level 1
Level 1

Try to config "snasw pdlog exception", this will filter out the informational log from the router log.

By configuring "snasw pdlog problem", it will filter out the informational and exception log from the router log.

But be careful with "snasw pdlog problem" in the normal situation, it might let you over look some problems. We strongly recommend that you don't use it normally.

Hope this help

Jing

romney
Cisco Employee
Cisco Employee

Jing, actually since the log message is type PROBLEM (most severe) there is no way to prevent it from coming out in the router log (snasw pdlog can only be used to filter out EXCEPTION or INFORMATIONAL log messages). There is also no way that I know of to suppress the DLUS event messages.

What I recommend is that you stop the link (via snasw stop command or nostart keyword on the snasw link configuration) and not start it until the switched major node is defined and active. You could also allow dynamic PU definition on VTAM by adding DYNPU=YES on the hpr-ip port ... then you don't have to worry about having your switched major node active.

Ray

To stop the SNASw event messages, you can disable console logging ("no logging console" in global config), or configure it with a level no greater than 2 (critical). A somewhat drastic measure ...

You can try to keep the upstream link inactive as Ray suggested if the DLUR_LOG_2 message affects all the downstream PUs. Otherwise, if only one or just a few PUs are causing these messages to appear, while other PUs are active and in session, stopping the upstream links would cause disruption !

Normally the SWNET member would be active. They IPL'd Saturday night and the member was not in the auto start list. It has since been added. When I came in Monday A.M. the log was full. It would be nice to have more control of which messages go to the log in SNASW. For example, EIGRP and OSPF have logging options.

What is the default SNASW PDLOG option?

romney
Cisco Employee
Cisco Employee

I'm not familiar with the logging options for EIGRP and OSPF ... I'll take a look and see if there is anything there that we should pattern after. However, the particular log message that you are getting is of the highest priority (PROBLEM), so the only additional control we could give you for snaswitch log messages is to filter them out entirely. The default snasw pdlog setting is to log PROBLEM and EXCEPTION messages and filter out INFORMATIONAL messages.

Based on your latest reply I have to apologize that I misunderstood which switched PU major node was not active. For some reason I thought it was the one containing the adjacent link station definition for snaswitch, hence my comment about dynpu on the hpr-ip port. I now see that you are talking about the switched major node with the dependent PUs/LUs in it.

romney
Cisco Employee
Cisco Employee

I do NOT recommend stopping the link as a workaround for the dependent PU/LU switch major node not being active. What you can consider doing is have a backup DLUS where the switched major node's PUs are in CONCT state and it would be used if/when the primary DLUS failed to activate the switch major node.

Ray, EIGRP & OSPF logging are much simpler with SNASw pdlog, it can only "log-adjacency-changes", it has 3 options, no log, log brief and log detail and there are only very few messages regarding the state of the neighbor. It's hard to compare EIGRP & OSPF logging with the SNASw pdlog.

Can you please help me to understand the following:

1. Are all the EVENT messages informational? I thought they are.

2. If they are, the following activation/deactivation messages should be filtered out from router log with "SNASw pdlog PROBLEM", is that correct?

SNASW-3-EVENT: Pipe to DLUS activated

SNASW-3-DLUR_LOG_2: PROBLEM

SNASW-3-EVENT: APPN to DLUS deactivated

Thanks

Jing

The SNASw EVENT messages are not pdlog messages; they are sent directly to the console log. They are quite limited in number.

tran

We need to make the distinction between the pdlog (snasw problem determination log) and the console log.

The default option for "snasw pdlog" is "problem": besides logging all pdlog messages to the pdlog, SNASw will send all PROBLEM category pdlog messages to the console log (if enabled).

However, there are EVENT type messages that SNASw will always send to the console log (if enabled).

In your case, you would need to disable console logging, but it may be impractical if the operator rely on these console messages.

A possible enhancement would be to add an option to SNASw not to send any message to the console log. But then you also risk of missing any really important warning from SNASw.

Thank you for your input and suggestions. There does not seem to be a good solution to this situation. I think adding an option to disable messages would be reasonable. If the link is broken the CISCO technician will be investigating the SNASW status and does not neccassarily need the messages in the log.

My experience with the CISCO log is that messages are usually written when interfaces or connections are made and broken. If more details are required, than deugging or additional logging options are available.

Regards,

Bill