| |
| VIRTUALISATION
BENEFITS |
|
RPO AND RTO |
|
Once your production systems are virtualised
disaster recovery becomes easy - just replicate
your virtual machines and bring-up the replica
copies either for compliance testing or for a
live fail-over.
Even use completely different hardware at the
recovery site if required.
|
|
When you start your
disaster recovery project it is essential that for
each application your organisation defines the:
- Recovery Point Objective - the acceptable
amount of data loss measured in time
- Recovery Time Objective - the duration of
time within which a business process must be
restored after a disaster
|
| |
|
|
| AUTOMATION |
|
STORAGE REQUIREMENTS |
|
VMware vCenter Site Recovery Manager automates
the process of:
- Performing a test fail-over to ensure compliance
- Performing a live fail-over in the event of
a disaster
- Performing a fail-back after a live fail-over
Without Site Recovery Manager complex manual
or scripted processes are required that may not
meet the required RTO.
|
|
It is often assumed that you can size the recovery
site array simply based on capacity, but this
is often an unwise strategy unless you are prepared
to accept a significant reduction in performance
in the event of a disaster.
For example 3TBs of RAID 5 300GB 15K drives provide
nearly seven times the random IO performance of
equivalent 1TB 7.2K SATA drives.
|
| |
|
|
| BANDWIDTH REQUIREMENTS |
|
IP ADDRESSING |
| As a rule of thumb
a dedicated 10Mbs link can replicate 80GB of daily
changes, introduce compression and optimisation
and this quantity increases considerably. |
|
Failing over one site to another requires either:
- Stretched VLANs
- IP address changes and DNS updates
|
| |
|
|
| REPLICATION TECHNOLOGIES |
|
SNS SERVICES |
|
There are many replication technologies available
today, it is therefore essential to determine
which one is right for your organisation. Choose
from:
- Asynchronous replication - a write is considered
complete as soon as local storage acknowledges
it, remote storage is then updated, but with
a lag
- Synchronous replication - guarantees "zero
data loss" by ensuring that the write either
completes on both sides or not at all, but overall
performance drops proportional to link latency
and distance
- Continuous replication - automatically captures
every version of the data that the user saves,
allowing the administrator to restore data from
any point-in-time
|
|
A successful disaster recovery project requires:
- Virtual infrastructure sizing for both sites
- Storage array sizing for both sites
- Bandwidth sizing to meet required RPOs
- Replication tiering to meet individual application
RPOs
- Testing and documenting fail-over and fail-back
processes to meet required RTOs
SNS can assist with all aspects of your disaster
recovery project to ensure a successful outcome.
|
Events
