03-06-2014 11:41 AM
I need to monitor daily if any port which were online got offline or having any issues ( in sync etc)
Since we dont have DCFM/Network Advisor yet, how can i enable this funtionality ?
i would like to avoid doing manual checks.
We dont have fabric watch either, can syslog or snmp help ?
What is the procedure to get this alerts ?
Any scripts ?
03-11-2014 01:08 AM
this is Fabric Watch or if you have FOS 7.2 Fabric Vision.
If you do not have those License, you have to Monitor your RAS-Logs manually.
That is the way live works. When you want to have comfort you have to pay for it
03-11-2014 11:35 AM
We have a script that gathers the errors for each port, saves them, then clears them. In that directory, I can grep the port in switchname* and see, for the last 90 days, whether it's been online, offline, and whether it's had errors. All using nothing more than ssh.
03-18-2014 01:04 PM - edited 03-18-2014 01:05 PM
You can use SNMP to monitor the status of ports to see if the went offline or not.
If you shop already runs Cacti or Nagios this is relative easy (1 day to get a POC going) to setup and cost you only a few hours work. Both Nagios and Cacti can alert when thresholds have been breached although may require an pluging)
The Names and OID you are looking for are (from Brocades SW-MIB)
swFCPortPhyState <> .184.108.40.206.4.1.15220.127.116.11.18.104.22.168.3
swFCPortOpStatus <> .22.214.171.124.4.1.15126.96.36.199.188.8.131.52.4
swFCPortAdmStatus <> .184.108.40.206.4.1.15220.127.116.11.18.104.22.168.5
There are similare ones from the FibreChannel Alliance FCMGMT-MIB
connUnitPortState <> .22.214.171.124.126.96.36.199.1.6
connUnitPortStatus <> .188.8.131.52.184.108.40.206.1.7
connUnitPortHWState <> .220.127.116.11.18.104.22.168.1.23
03-21-2014 02:25 PM
Thanks for the post, its very usefull, it will get what i want , however, my intension was not to keep customizing each and every alerts and have to code something new for things i did not think about, i.e FCIP tunnel going bad, BB credit deplition, port errors increasing, i would have to come up with a plan for each of this seperately, does BNA offer built in capabilities which makes things easier and catch most of the abnormalities ?
Does third party tools such as orion any good ? is fabric watch needed to send snmp alerts of interest, i know you can alert snmp natively but then why Fabric Watch, what addtional capabilites does it bring?
lots of questions , not enough answers.
03-23-2014 03:34 AM
standar SNMP traps will notify issues such as a defective SFP, FAN, CP, etc. Fabric Watch allows you to configure snmp traps/emails for a way bigger number of issues: if a port reaches a certain amount of CRC errors, or a certain number of frame discards; if CPU is above a determinate threshold, etc.
I personally find it very useful for t'shooting.
03-24-2014 08:27 AM
AFAIK Orion uses SNMP as well, if it has pre defined templates to do what you'r interested in , by all means go for it.
Unfortunately I don't know Orion at all. I use SNMP for everything i need combined with tailored templates in Cacti.