I have the question about the effectiveness of regular rebooting of Catalyst 6500.
I personally think that regular rebooting will take effect of cleaning up fragmented memory which might be caused memory leak problem.
However regular rebooting will impose to schedule of network down, arrange person, adjust business with external company and risk of the causing problem at boot up.
So I also personally think if Catalyst 6500 operates with stability, there is no need for regular rebooting.
Do you execute regular rebooting ?
Do you recommend regular rebooting ? if so, why ?
Your information would be greatly appreciated.
We work on the principle that "if it ain't broke don't fix it" ie. we do not do regular scheduled reboots of our 6500 switches.
A lot depends on the environment you are running in. Where i work we have so many interested parties in terms of applications in our data centres it is very difficult to get downtime to reboot the switches. So it really is a last resort. Obviously in a fully redundant setup ie. teamed NIC's, access/distribution/core paired switches etc. we could take one down without disrupting service but even this is a major organisational issue.
I have also seen problems when rebooting where perfectly functioning modules upon reboot suddenly decide not to work and then you have just created a problem that you really didn't need.
If however i worked in a 9 to 5 environment and anything outside of those hours was not needed then i admit i would be tempted to schedule a reboot every now and then. But there are very few 9 to 5 environments these days !.
Also bear in mind that one of the things you are paying for when purchasing a 6500 is the ability to dynamically insert and remove modules without rebooting so these switches really are meant to be up and running most of the time.