Turns out the Small Form-Factor Pluggable Gigabit Interface Converter (SFP GBIC) has a vey limited number of insertions in and out of a chassis. A range of 100 to 500 is specified for older GBICs but for newer types the lifetime rating for insertions on the Nexus 7000 is between 150 and 200 cycles (of insertion).
Has a big impact on using modules in lab and testing environments especially where things might be changed very often. 150 insertions for a GBIC is not that many lab setups.
Reference.
Cisco Documentation
Unnecessary removal and insertion of a GBIC could lead to premature failure of the GBIC. A GBIC has a lifetime of 100 to 500 removals and insertions.
Molex Connector Production Specifications for Small Form-factor Pluggable (SFP) Stacked 2-by-6 Multi-Port Connector with 4 Lightpipes, with Press-fit PC Tails, 240 Circuits – These are typical examples of the connectors used in the manufacture of switch modules for just about every networking vendor.
Durability (mating cycles max) 200
http://www.molex.com/datasheets/pdf/en-us/0754510001_IO_CONNECTORS.pdf






This issue is not helped by the often poor manufacturing quality of the SFP ports on Cisco switches (the holes are often mis-sized or just plain wonky).
The GLC-T is also incredibly easy to break if installed/removed incorrectly.
Greg.
on a similar note.
someone in my organization was testing a SW/FW upgrade procedure on a nortel switch. After his testing was complete this person reported back to me that the upgrade procedure was flawed and the upgrade process corrupted the switch and left it inoperable.
I placed some calls at Nortel before opening a ticket.
I later found that this person was performing extensive testing. Extreme stress testing if you ask me. He upgraded the SW and FW 30 times in a row on the same switch. Each time, initializing the config, and downgrading the switch to a lower version before starting again.
After a small laugh, Nortel replaced the switch. But with the kind reminder that most switches will have only 5 major and 10 minor releases during the life of the switch. The NVram was never tested to be writen to over 60+ times.
everything has a breaking point.
Hi Paul,
There was/is an known issue with some of the Nortel/Avaya 5000 series stackable switches that was only recently resolved where the NVRAM/FLASH would experience a failure due a high number of configuration writes when MAC security was enabled in the configuration.
I made a post about it on my blog.
Cheers!
Hey Mike, that is interesting. I was not aware it was a known issue. In this case it was a 470. But I am sure allot of the switch bits and pieces are mostly the same
Great article. re-blogged it to mine.