Re: Checkpoint_segments optimal value

On 7/25/2014 5:58 PM, Prabhjot Sheena wrote:
> We do have i/o problem during heavy loads because we share mount point
> from netapp across different machines which we are getting rid in few
> weeks by moving stuff to Amazon aws and than IO issues will go away
> with fast iops

Netapp has to be some of the most expensive storage I've ever seen in a
datacenter and is frequently a source of performance problems.   NFS
isn't really appropriate for storing database data on, there's way too
many opportunites for fsync to get lost, more so if you've got a
virtualization layer in there too.





--
john r pierce                                      37N 122W
somewhere on the middle of the left coast