This post is more than 5 years old
197 Posts
0
2275
December 19th, 2014 05:00
RecoverPoint "Maintenance Mode"
Does RecoverPoint have a "maintenance mode" feature similar to ESX hosts/vcenter where all the CGs can be slowly moved to other RPAs? When we reboot a RPA it causes our VNX CPUs to spike to around 90%, which I assume is due to all the CGs entering the init process.
Image below is the impact of rebooting 1 RPA at the DR side. (DR VNX is green and black)
No Events found!
forshr
2 Intern
•
1.1K Posts
0
December 19th, 2014 08:00
RecoverPoint does have a maintenance mode but this is used to prevent change to the RP configuration and not in the same way as ESX maintenance mode. What you are seeing is the impact of the data being read from the splitter and/or array due to the transfer of the CGs, i.e. a short init. If this is causing an tangible effect on host latency then you can be throttle the reads back using the config_io_throttling CLI command. Obviously the slower the RPAs can read, the longer it takes for a CG to reach an Active state after it has been transferred.