09-20-2011 06:19 AM
Can you help me, please?
I have a fabric with Brocade B40 switchs.
I see a lot of enc-out and disc-c3 errors at ports.
frames enc crc crc too too bad enc disc link loss loss frjt fbsy
tx rx in err g_eof shrt long eof out c3 fail sync sig
0: 377.1k 563.9k 0 0 0 0 0 0 0 0 0 0 0 0 0
1: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2: 357.7m 151.0m 0 0 0 0 0 0 0 0 0 0 0 0 0
9: 2.1g 1.3g 0 0 0 0 0 0 0 20 0 0 0 0 0
10: 3.7g 16.2m 0 0 0 0 0 0 2.3m 0 1 2 1 0 0
11: 4.2g 2.2g 0 0 0 0 0 0 0 0 1 2 1 0 0
12: 3.1g 1.0g 0 0 0 0 0 0 0 138 0 0 0 0 0
How do i see date and time of errors ?
Fabric does not notifiy me when an error occurs at port link.
The fabric version 5.3 has fabric events. At fabric events i can see failed port link with date and time.
The fabric version 6.4 does not have fabric events.
Where do i see failed port link at fabric version 6 ?
09-20-2011 07:19 AM
As the prompt (IBM_B40_...) suggests you most probably have an IBM branded Brocade 5100. IBM ships them with the Fabric Watch license. Fabric Watch is a powerful toolset for monitoring ports and the switch overall. Please have a look into your documentation, (which you can also download here at brocade). There should be a Fabric Watch Guide which explains how to configure it to get the needed alerts.
09-20-2011 08:21 AM
Yes fabric watch is a good tool. Another way that you poll via SNMP the error counter of intrest and let cacti or MRTG create some graphs.
A third way you create some scripts which collects the error count from the switch. Run these scripts in a specific time interval. But this will give you only a rough idea when it happens.
It depends almost on your SAN size and the amount of time you would like to invest.
All three ways have its own advantages or disadvantages.
I hope this helps,
09-22-2011 03:04 AM
Thank you very much for all the answers.
I haven´t experience with SNMP.
I liked the idea of using Fabric Watch. I'm setting it up.
But Test Email Alert doesn’t work.
Why doesn’t it work ?
09-22-2011 06:42 AM
Sorry I can't see the second picture.
But have you setup the DNS correctly?
Try first the IP instead of the FQDN.
Check the mailserver if it bounce the mails
By the way try to configure F/FL Port and E ports and not the Ports elements.
In the past I had issues with them and I use only F/FL and E ports without any issues.
09-23-2011 11:09 AM
what value you have setup for fwalarmsfilterset ? Also check `dnsconfig` look at `troubleshoot fabric watch using email alerts` on fabric watch guide. http://download.oracle.com/docs/cd/E19055-01/sw.12000/CRCM1519/CRCM1519.pdf
09-25-2011 01:51 PM
what does it mean to say by "configure mail server on the switch and enable alert for the class"? after reading the pdf it looks to me we need to enable fwalarmsfilterset. Do you have an environment to test it?
09-28-2011 03:10 AM
email notification is working.
The problem was at dnsconfig.
I’m very happy.
Now I have to wait for events to analyze notifications.
I have another question.
Where can I get documentation of construction of command scripts for maintenance fabric ?
My Fabric has 200 ports. I don’t want to clean the statistics one by one.