White Papers

BP1013 Best Practices for Enhancing Microsoft Exchange Server 2010 Data Protection and Availability 22
Figure 7 Time taken for 1 to 8 Smart Copy snapshots of one volume under a load of 5000 concurrent users
Figure 8 Time taken for 1 to 8 Smart Copy snapshots of one volume under a load of 5000 concurrent users
with one active host only
These results clearly show that the process of creating a Smart Copy snapshot had a predictable and
minimal duration, with an average time of less than 5 seconds. Also, and most important, the trend did
not change with different workload conditions, even when a critical situation was simulated. A minor
trend is that the snapshot duration increased slightly when the user load was increased and when the
database count running on the same host was increased. This is due to the increased effort required
for freezing the I/O in transit which is higher when the user load increases.
6.2 Host resources impact
When collecting performance indicators from a mailbox database server, the focus is generally on
storage response time because it has a direct impact on the user experience, but other traditional
indicators (processor, memory, and network) should be kept under surveillance as well to ensure
overall monitoring of the environment.
With the profile of the users created as a fixed value (Outlook online, Heavy), the increase in storage
access from a mailbox server progressed with the amount of workload applied. We retrieved some of
the Exchange key performance indicators during all the tests, but the most significant indicators for
our analysis were retrieved during the maximum workload.