Graduates course- Security

  • Thread starter Thread starter Taishi
  • Start date Start date
T

Taishi

3 Windows 2000 servers

I am setting up a E-commerce project for my Graduates course called
Security. It's a small business with servers. I want to make sure I am
wording this correctly and that I am technically correct.

3 servers setup for redundancy and failover

I think in the real world it should say "2 servers setup for redundancy and
failover". Because if server1 fails then server2 will have a mirror
image(replicor) of server1. It will automatically go into production when
Server1 fails. And for some reason, I don't see having a Server3 nor
Server4 for failover. Seeking expert advise.

Regards,
T
 
Hi,

You need to choose your words carefully.

Redundant Cluster:
This would mean that you setup more than one server and
they have a heartbeat running across them. Both servers
would share the load and in case of a failure, the other
server would know about the same thru the heartbeat.

Failover:
I define them of two types

Hot failover:
This would typically mean that I have a spare server
(hardware) either configured same as the production server
or one in which I can hot swap the hard disk. On failure
of the main server, I can quickly get started with the
failover requirement.

Cold failover:
This would mean typically having a standby low capacity
hardware to take care of emergency situations. Either hot
swap the hard disk of the failed server or load the app
again and get started. A stopgap measure and the downtime
is maximum amongst the three mentioned.

Hope this helps..
 
3 Windows 2000 servers

I am setting up a E-commerce project for my Graduates course called
Security. It's a small business with servers. I want to make sure I am
wording this correctly and that I am technically correct.

3 servers setup for redundancy and failover

I think in the real world it should say "2 servers setup for redundancy and
failover". Because if server1 fails then server2 will have a mirror
image(replicor) of server1. It will automatically go into production when
Server1 fails. And for some reason, I don't see having a Server3 nor
Server4 for failover. Seeking expert advise.
Click to expand...

Failover, or redundancy failover is normally two systems. Clustering
is two or more. In Failover, the backup system is idle until the
primary fails, in a cluster all can be active at all times.
Clustering provides both redundancy as well as performance increases.

Neither actually directly relates to security, though depending on
your graduate course that fact may be lost on the faculty... :)

Jeff
 
Clustering provides both redundancy as well as performance increases.
Click to expand...

I have to disagree with that last statement. The clusters I have worked
with have all suffered performance hits due to sharing storage. Most
experts suggest that clustering be used strictly for high-availability, not
for performance, expecially if you are using SQL.
 
Back
Top