OK, here comes my 18.104.22.168 report:
After updating to 22.214.171.124 all of our 8164´s (8 switches) report HIGH temperature alarms in the web gui. But the reading shows 45 degrees celsius - nothing huge as far i can say.
When power is lost very short on a unit (but the other stays OK) the fan will report failure forever (even if power is restored). This can not be cured with a reset, you have to disconnect the whole switch from power for about 10 secs.
Somewhere in the past you adjusted the intel x520 timing so that it would no more stay offline when powercycled. Well, i suggest you re check this after 126.96.36.199 again please - i saw some strange behaviour this morning with my x520s (had to re-plug em in to get em recognized).
Especially about 1) has to be taken care of ASAP. All my switches are in alarm state now. The datacenter where they reside has totally perfect temperature of 20 degrees celsius. When the switch reads 45 degrees on the mac board i thought that should be perfectly normal. I really hope this is not connected in any way to the cpu change to revision b you did the last weeks. All my 8164s are revision a. Please clarify ASAP, this is very important for us.
The switches are fully equipped with 10gb sfp+ modules and also the 2*40 gig optional module for each one.
Thank you for bringing these concerns to our attention. With new firmware and potential fixes that need to be addressed, the best path to resolution is to call in and have a case opened. The case can document everything and proceed with a path of escalation if needed. The number to phone support is
ok, thanks, if i find some time in the next weeks i certainly will open cases for 2 and 3, but for the moment could you please forward number 1 to the dev team? I documented it with my picture and text and there is not really much more to say about it - it´s just as it is and the picture speaks for itself.