How about teamed NICs going to different physical switches? That would
address the specific fault you experienced.
Also, some kind of monitoring and alerting for service down would help.
--
Geoff N. Hiten
Senior SQL Infrastructure Consultant
Microsoft SQL Server MVP
"LanLan" <LanLan (AT) discussions (DOT) microsoft.com> wrote
Quote:
Basically, we have a 24/7 database on the MS SQL 2005 cluster. Last week,
the
cluster failed due to the NIC fault. The service was stopped for almost 4
hours. The penality is heavy. My manager want to add another layer to
ensure
the availability of service, such as data mirroring.
I just want to check whether anyone here has adapted such technology and
could share some experience. |