Disaster recovery solutions:
Issue
What will it cost the company if the server is down
for an extended period?
What will happen to your business if the server gets
stolen/damaged or has a major hardware failure?
Solution
1. Must have Raid on server and centralise all data.
2. Daily tape backup and keep offsite.
3. Prepare a backup server.
4. Create images of all your servers and keep them up
to date.
6. Secondary backups onto NAS or other disk storage
7. A spare backup server offsite.
8. Service agreements on your server hardware.
9. Routine maintenance on your network equipment
10. Type a disaster recovery plan.
Benefits
- Data security
- Less down-time
- Insurance on major disasters
- Peace of mind.
Disaster Recovery Planning
With the level of dependence many organisations place
on their information systems, it is necessary to consider
the business consequences of a potential system failure.
- For example what is the financial cost to your organisation
for a single day of lost productivity due to an unexpected
system failure?
What would happen if……
• A new ‘Virus’ or
‘Worm’ infects your system
and corrupts the system rendering it
unusable??
• A power surge damages critical
server components??
• A theft occurs and your server
is stolen?? (It’s happened before!)
|
|
|
System failure commonly occurs due to one or more of
the following reasons;
- user error, - inadequate maintenance, - hardware failure,
- software conflict, - or as the result of a deliberate
attack.
Unfortunately there is no guaranteed method to protect
against all the above issues. Therefore the best course
of action is to be aware of the potential risks, take
precautions wherever possible and be prepared in the
event of a failure.
All-Tasks employ a two-part strategy to contend with
system failures. (Hardware or software.) Firstly we
offer comprehensive preventative maintenance to actively
manage the network infrastructure. Secondly we work
with our clients to implement procedures and put in
place tools that allow us to rapidly recover in the
event of a systems failure.
Should a ‘worst case scenario’ occur, the
procedures and associated tools will already be in place
to get the system back up and running in predefined
time period. Common expectations are for a 4-hour or
an 8-hour recovery window.
Without forward planning it could take 15 – 30
hours to rebuild a downed server..
Disaster Recovery Implementaion
The single most effective tool in ‘disaster recovery’
is having an image of the working system volume of your
server. (This image will include your specific program
and user configurations.) Should a critical failure
occur, rather than re-build the server (installation)
from scratch we can restore a known working image.
Q.
I have a tape backup, isn’t that
doing the same thing?
A. Unfortunately not.
Tape backup is an archival media primarily
used for data storage. You cannot effectively
backup a working system volume (programs)
with a tape backup, it’s designed
to store data.
A tape backup device can potentially
take hours to backup a system and by
the time the backup is complete many
things may have changed. Secondly you
cannot use a tape backup to restore
data until the system is back up and
running. If the system is corrupted,
you will need to rebuild it first. |
|
|
There are many methods of implementing system imaging
technology. The simplest version is an off-line image.
Once the server is working in a satisfactory state,
we take the server offline. While the server is shutdown,
we take a complete copy of the working system and store
it on another hard drive. Once the image is complete
the backup drive is removed from the working system.
The whole procedure should take under and hour and will
enable us to have 15+ hours work in the event of a system
corruption. The cost to the client is minimal, the hardware
is one additional hard drive.
The downside of off-line imaging is that it is a manual
process and the image needs to be regularly updated
manually, which involves taking the server off-line.
Ideally the image should be updated every 3 to 6 months,
or anytime significant changes are made to the configuration.
(Especially when rolling out updated applications.)
As a further benefit, the drive also becomes a hot-spare.
Should one of the live drives fail, we have a spare
on hand to replace it with.
Taking it further
There are better methods of server imaging available.
The ultimate solution is to use an advanced software
tool to automatically create a daily live backup on
another server. The image can then be archived to tape
for storage.
|
