Brocade DCX monitoring with virtual fabrics

Michael.Kirschig

I try to monitor our SAN infrastructure with stor2rrd. Each fabric consist of 2 DCX8510-8 and some Blade switches. When I add the DCX with snmpv3 (because of some virtual fabrics) I get lot of events: /home/lpar2rrd/stor2rrd/bin/san_stor_load.sh: 2016-08-25_13:17 : too many defnfc0061 processes: sanperf.pl, exiting

and the snmp daemon on the DCX is about 60-100% cpu.

any ideas?
Michael

Pavel

hi,

chnage cron schedulle from every minute to 5 minutes (or 10 minutes if 5 minutes still causes a problem)
Some old firmwares have a problem with every minute providing stats through snmp v3.

0,5,10,15,20,25,30,35,40,45,50,55 * * * * /home/stor2rrd/stor2rrd/load_sanperf.sh >/home/stor2rrd/stor2rrd/logs/load_sanperf.out 2>&1

Michael.Kirschig

Hi,

actually I have one fabric with 7.3.0c and the other with 7.4.1d. Even with 15 minutes polls the same Problem, but only at the DCX Switches!

 

Pavel

Hi,

I affraide we cannot do too much here if the switch is not able to provide data every 15min withour overloading itself.
You can ask your SAN switch support if this is a bug in your firmware etc.