Cisco router 3900 IOS Version 15.0(1)M2 restart when run tcl script

[duongthanhphao1]

When i run tcl script command " scripting tcl init flash0:bvr_pe01.txt "

after i run command " tclsh " router restart

I don't know why ?

File script tcl as attach below

Please view and reply! Now i am confuse about it

[Mark Malone]

Are you doing it by SSH

CSCtt21228

Symptoms: Router crashes while trying to configure Tcl script via SSH connection.

Conditions: SSH to the router and then try to configure Tcl script.

Workaround: There is no workaround.

Looking through the release notes there's a lot of TCL bugs in that version

http://www.cisco.com/c/en/us/td/docs/ios/15_0/release/notes/150MCAVS.html

[duongthanhphao1]

Thanks so much mark malone but I am doing it by telnet to run tcl command

[duongthanhphao1]

Please help me no one reply

[Vinit Jain]

Hello,

I believe Mark already pointed to one of the software defects. Though I am not really sure on the IOS version that you are using:

Could you please share the below information:

- show version

- Crashinfo file (might have been generated when the crash occurred)

- Console logs during the time of the crash

Once I have the information, can look into it to see what is the cause of the crash and how it can be fixed.

Rgds,

Vinit

[duongthanhphao1]

Thanks so much Vinit Jain

I send log files as your recommand

[Vinit Jain]

Thanks for the logs. I would like to clarify one thing here. If you remove the "repeat 1000" command from the tcl script, do you still see the problem.

The crash is happening on the ICMP event handler process:

---- Partial decode of process block ----

Pid 128: Process "ICMP event handler "
stack 0x136B9BF4 savedsp 0x10A60C4 
Flags: analyze prefers_new 

Just wondering if you have checked the above. Also, would like to know if you just have 10 host addresses to ping, does that make any difference.

Regards,

Vinit

[duongthanhphao1]

Thanks for your reply quickly.

When i use the same IP ping list and use repeat 100, i don't have this issue

[Vinit Jain]

Thanks for the information. Any reason for having the repeat count to 1000 or 100. If you want to check the reachability, you can use it without repeat counter or you can use IP SLA.

Also, 15.2 release is a more stable release in case you want to overcome these issues.

Hope this helps. Please let me know if you have any further questions.

Vinit

PS: Please rate all the posts if you find the information useful and do mark the question as answered if your query has been answered.

[duongthanhphao1]

Thanks for your reply, expecial thanks for your recommand using IP SLA changing using IP ping but when using IP SLA, i don't know how my staff that is monitoring to know when is not reachability ?

You can explain more detailly by using IP SLA

[Vinit Jain]

IP SLA can be used with ICMP to check the reachability of the hosts. You can also integrate it with EEM script which can automatically send emails when a reachability has failed.

One quick example of IP SLA is as below:

http://blog.codergenie.com/blog/post/2015/02/04/ip-sla-with-tracking.aspx

Hope this helps.

Vinit

[duongthanhphao1]

Thanks for your reply. You can send me document about EEM script which can automatically send emails when a reachability has failed and detail example ?

Thank you so much!

[Vinit Jain]

Refer to the below post :

Email alert on IP SLA based packet loss:

https://supportforums.cisco.com/discussion/12225606/eem-email-alert-ip-sla-based-packet-loss

Hope this helps.

Vinit

[duongthanhphao1]

Thanks for your support

I am still affaird that i must monitor many ips and do it affect to performance Router ? I can only create one SLA for all of IPS monitor ?

[Vinit Jain]

Though you will have to configure multiple IP SLA probes but its still better than the router crash that you are having due to TCLsh. you can still go for tclsh but i till not really recommend to perform 100 or 1000 pings from the shell. Its better of checking 5 pings and based on that you can figure out if the host is reachable or not and also, it will be faster process.

I dont think you can have same SLA for probing multiple IP's. You'll have to configure multiple probes.