Video Screencast Help

So how long does post EV10 upgrade indexing take? Some numbers.

Created: 15 Oct 2012 • Updated: 15 Nov 2012 | 2 comments
DeadEyedJacks's picture
This issue has been solved. See solution.

Here's a real world example which might help your server sizing / time estimates.

Site has two EV storage / index servers plus dedicated SQL server, all now VMs under vSphere 5 recently migrated from old physical servers. Due to legacy reasons journaling and mailboxes archives have shared storage partitions and SIS parts.

EV servers have 8 vCPUs and 16GB RAM, SQL has 4 vCPUs and 8GB RAM, each on seperate ESX host. ESX hosts are dual socket, six cores, 2.7Ghz, 96GB RAM, 10GB Ethernet, Jumbo Frames, seperate LAN, NAS and SAN fabrics.

Indexes and SQL database on SAN storage,  Vault store partitions on NAS CIFS storage.

Index and database writes produce 600 IOPS, partition reads produce 1200 IOPS,

After one week of continously running index upgrades, 15 million items have been processed.

Hope this proves useful.

Comments 2 CommentsJump to latest comment

AndrewB's picture

how is your SAN and NAS storage attached to the VM guests?

Andy Becker | Authorized Symantec Consultant | Trace3 | Symantec National Partner |

DeadEyedJacks's picture

SAN storage is presented as guest iSCSI to VM. SAN has its own VMware extensions for load balanced multi-pathing.  NAS storage is CIFS remote share.

Authorised Symantec Consultant on Archiving and eDiscovery ASC, STS, SCS, SSE+

Microsoft, NetApp and VMware certified professional MCTS, MCSE, MCSA, NCDA, NCIE-BR, VCP, VTSP