Performance,
Stability, Scalability
All three of products which together comprise the Isode Service Provider
offering, return performance figures unmatched in their class whilst
demonstrating the robustness and security required in an internet messaging
environment.
M-Switch
M-Switch Anti-Spam was benchmarked against the open source product
SpamAssassin (popular with many Service Providers with high-volume spam
processing requirements) and found to be significantly superior in both
message processing speed and spam identification accuracy.
When processing messages M-Switch Anti-Spam (using the included SMTP
message switching) was over 700% faster than than SpamAssassin (using
Postfix as the message switch).
M-Switch Anti-Spam intercepted in excess of 95% of spam with a false
positive rate of 0.15%. M-Switch Anti-Spam, when combined with the message
quarantine features described here, ensures
that spam is effectively filtered whilst no vital business messages
are lost. Details on how the benchmarking was conducted and the results
can be found here.
M-Box
In benchmarks described in this
whitepaper, M-Box was compared to a range of leading competitive
IMAP servers using MStone, a freely available industry standard stress
and performance measurement tool.
M-Box significantly outperformed competing products in terms of both
transactions per second and errors generated. The differences were particularly
marked at higher load levels, demonstrating the excellent scalability
of M-Box.
M-Vault
In tests against OpenLDAP using the independent DirectoryMark test
framework, read and search functions were assessed as well writing and
loading mixed tests combining a variety of operations to simulate a
general purpose directory operational environment.
M-Vault consistently
outperformed OpenLDAP across all tests for database sizes ranging
from 50,000 entries 1 million entries.
Conduct your own tests
All Isode products are available for evaluation
and in all of our benchmark whitepapers we have taken care to use widely
available test software run on reasonable hardware platforms to encourage
evaluators to repeat and verify our results.