Microsoft Corp. said Monday that its upcoming Exchange Server 2010 software will be able to read and write e-mails to disk 70 per cent faster than Exchange 2007, despite the fact it uses the same database engine.
Last month, Microsoft released the public beta of Exchange 2010, which is due by year’s end. At the time, it unveiled a number of flashy end-user features designed to match or keep pace with Google Inc.’s Gmail.
At the Tech Ed conference in Los Angeles, a Microsoft developer detailed a handful of back-end improvements in Exchange Server 2010.
All address performance and scalablity, which has long been a sore point for Exchange administrators and users, with Exchange’s Jet database engine taking much of the blame.
Several years ago, Microsoft announced a plan to move Exchange to a SQL Server-based database engine, but later abandoned it.
Better, greater storage
While keeping Jet for Exchange 2010, Microsoft changed its storage schema such that reading and writing mails to disk will be smoother and less bursty, according to Ross Smith IV, a senior technology architect for Microsoft.
That will result in a 70 per cent I/O improvement over Exchange 2007, and 90 per cent improvement over Exchange 2003, he said.
Exchange 2010 mailboxes will also be able to store up to 100,000 items per folder, up from 20,000 in Exchange 2007 and 5,000 in 2003, he said.
Exchange 2010 is even more role-based than prior versions.
In Exchange 2003, servers running the software could be configured as either a front-end proxy or a mailbox server, which offloaded work from the core mailbox server.
In Exchange 2007, servers can be configured as one of 5 roles: hub transport, mailbox, unified messaging, edge transport and client access.
Exchange 2010 still has five roles.
However, client access servers are being “beefed up,” such that they become a true middle-tier framework that handles all data rendering and conversion for Outlook as well as Outlook Web Access, said Smith.
30 second failover
That will free up a mailbox server in Exchange 2010 to handle up to 120,000 active connections, versus just 12,000 or so in Exchange 2007, he said.
Mailbox servers will also now be able to fail over to another one within 30 seconds, compared to two minutes or more for Exchange Server 2007.
Another new feature called shadow redundancy will enable two copies of an e-mail message to be kept so that administrators do not need to run a RAID backup.
Some smaller companies and branch offices may run all of their Exchange 2010 roles on a single server with up to 16 processor cores. But most should run multiple servers, for safety reasons, he said.
“We would much rather see you scale out than scale up,” Smith said.
Servers performing client access duties need at least eight CPU cores and 8GB of RAM in Exchange 2010. Servers can have up to 12 cores, but more may lead slower performance due to “processor cross-talk,” he said.
The biggest change for administrators in Exchange 2010 is an extended but simplified capability to distribute different user mailboxes across different servers, and keep those servers highly available.
The binding of message stores to physical servers has been loosened considerably, and network managers should now easily be able to replicate a message store — up to 16 copies are supported — across multiple servers, with automatic failover capability between servers.
Of course, having two copies of the message store doesn’t help if you don’t build other resiliency, such as multiple client access servers, into your deployment, so this new feature isn’t going to be the last word in simplified reliability.
Clustering is an important part of a high availability offering as well, and Exchange 2010 should simplify that tremendously.
In Exchange 2007, Windows Server clustering was managed very separately from Exchange itself, which required additional expertise and a different skill set from what the standard e-mail manager holds.
Exchange 2010 doesn’t entirely solve this (based on the documentation we received), but does move cluster management directly into the Exchange management system.
Some e-mail managers are reluctant to use clustering because they don’t understand it, or don’t know how to manage and control it; by moving this important part of a high-reliability system directly into Exchange, Microsoft makes it more likely that managers will be able to use clustering successfully.
An improvement on the scalability front is the ability to move a mailbox between databases without shutting it down.
In Exchange 2007, moving message store mailboxes from one database to another could require a significant amount of downtime.
In Exchange 2010, you move a mailbox which is in active use between message store databases.
This lets the e-mail manager balance the load across servers and disk subsystems without making mailboxes unavailable.
Having this feature will also let e-mail managers resist the temptation to build many small message databases rather than a few larger ones, because there’s no need to try and predict how big each mailbox and message store database is going to be for load-balancing purposes.
Exchange 2010 is also internally more resilient to failures, with the ability to automatically route around and retransmit messages lost by a malfunctioning transport hub.