Hi,

I am starting to write and learn about redundant systems.

Iam try to make a redundant system from a big system in the company that it does not have any redundancy right now.

There are a few concepts I dont understand.

Systems and technologies like:
- HA,
- Heartbeat,
- Pacemaker,
- DRBD,
- iETD,
- AoE,
- ESX,
- LVM,
- iSCSI,

--> can all this work together?
--> are there any reviews what is better?
--> are there many concepts to choose from (to bild an redundant server system)?
--> ?


I would like someone to tell me the basics and a few comments that will guide me what to search so I can understand better.



Thanks in advance

As you mention, there are a lot of way to setup redundant systems.
Active/Passive
Active/Active
Clustering

Are you trying to achieve redundancy or performance? Since your question is too general, I will suggest you should start reading below tutorials.
http://www.howtoforge.net/howtos/high-availability

Cheers

Hi, thanks for answer,

I am trying to understand a little bit more.

I know a little bit what these components means: High Awaliability (system design), DRBD (replicate data betwwen servers), Heartbeat (checks servers if they are still alive), Pacemaker (almost like Heartbeat), OCFS 2 (cluster file system from Oracle), iSCSI (network transport...), LVM (logical volume manager), AoE (ATA over ethernet, standard base protocol), ESXi (VMVare virtual base servers).

Ok I understand that we can set up how will system be redundant,
is it going to be active/active, active/pasive,

clustering: does it means that data is gruped in clusters, so it checks if all data is clustered the same on all servers?

Still I can not figured it out how system must be set for perfomrance or redundant.


I am searhing on internet if there is any review what is better?



Thanks againg for answers

There are categories
Cluster Filesystems (below will replicate local disks each other)
DRBD (replicate data at Block level)
OCFS2
GFS

External Storage communiate via Fibra channel or network protocol
iSCSI (basically you are using storage array as your local disk)
AoE (same logic but more lighweight)


LVM (filesystem nothing to do with HA)
ESXi (VMware Virtualization nothing to do with HA)

If you see above technology, here is where data store
DRBD - both machines has identical data
OCFS2, GFS - every machines has identical data
iSCSI, AOE, SAN - apparently data is stored at outside storage

What is better?
It is really depends on your requirements.

1 members found this post helpful.