UNSOLVED
M
michael_church1
1 Message
0
1652
February 13th, 2013 04:00
VNX replicator Performance issues
All I am working on a customer site with two Unified 5300 Boxes. I have setup VNX replicator to replicated CIFs file system between sites over two 1 Gbps links. However one of the file system just will not catch up. This contains User data and a large number of PST which are believed to be the issue. Is there recommended setting for the time out os sync or way i can increase the netwrok usage on replicator. As I am struggling to find a way of making the VNX use it available bandwidth. Normal issue networks say it is ok and EMC only suggestion is to changed the time out of sync to the default. At present it is set to 30 minutes I also looked at the server_netstat -info command and it looked to show that the physical ports were set to use an mtu of 9000 but the virtual interfaces used 1500 as there mtu. Is this correct in terms of setup
christopher_ime
6 Operator
•
1962 Posts
546
1
Posted July 9th, 2013 00:00
emc59083: "server_netstat -i does show wrong MTU size for ace Gigabit Interface"
christopher_ime
6 Operator
•
1962 Posts
546
0
Posted February 16th, 2013 17:00
Michael,
Firstly welcome to the forums and above all thank you for being an EMC customer!
Consider it the maximum supported MTU as viewed from server_netstat. The command will always report 9000. The value defined when creating the interface is what is applied. An MTU of 1500, which you have set, is generally the preferred setting over the WAN. Since latency and packet loss is bigger concern over what can be a relatively unstable network, packet retransmissions for a larger packet would be costly.
So are you not getting the throughput/bandwidth that you would expect? Just curious, was there anything EMC (support?) may have reviewed before they came to the conclusion of lowering the max_time_out_of_sync setting?