Some time back I wasted 5 days trying to find out why my snapmirrors were not keeping up. Netapp eventually very kindly told me about the global snapmirror throttle which I immediately disabled on all nodes and snapmirror speeds went up by a factor of 10.
Life was sweet.
Till this week.
Again my snapmirrors could not keep up. But I knew it could not be the global snapmirror throttle because I had disabled that before.
It had to be network right? We had made some network changes.
After 2 days I decided the symptoms were suffiently similar for me to revisit the global snapmirror throttle, and yes, sure enough the settings had reverted to enabled.
I suspect this is because we power cycled our netapp heads as part of a DR exercise. It looks like when the head comes up the setting to disable the global snapmirror throttle is lost.
Great stuff.
So, its up to a total of 7 days lost because of the global snapmirror throttle which as I said before seems to exist solely for the purpose of making sure things do not work properly.
Sorry, I had to vent.
Regards, pdg