Unsolved
This post is more than 5 years old
46 Posts
0
563
How do I determine if RecoverPoint is write constrained?
I have an 8 node CDP cluster 3.5 sp2 with 8 distributed CGs. I am protecting a 150TB SybaseIQ database.
Almost constantly several of the CG's are in a highload state.
Yet when I look at throughput and IO/s I see avg. of 30MB/s and 500 IO/s per PRA which is well within the published limits of 120MB/s and 20000 IO/s. System traffic peaks at 8000Mb/s on occasion. No journal compression to save CPU cycles.
CPU usage on the individual RPA is approx 200% with RPA1 constantly at 400%.
The storage is all Clariion or VNX and I do have the maximum amount of splitters attached.
I have run balance_load a couple of times over the last month to see if it can smooth out the load but its not made any difference.
How do I determine if the journals or target storage are write constrained - there must be a queue I can look at somewhere?
Is there anything I am missing?