Link to home
Start Free TrialLog in
Avatar of yodaa
yodaa

asked on

Disaster Recovery & Business Continutiy Plan

Hi All

Im planning to create a plan for BCP for my network, can someone advise any solution.

My plan is:

1. Create Workstation  and install Server 2012 Host then from the backups install all virtual servers on it, create new RDC and point to different public IP.
2. Create one WKS with multi seats software to be able access remotely  for all staff.
3. Put these machine in different location/place.
3. In disaster all member will be able to connect remotely from home BCP Machines.

What do you think ?
Any Advice ?  
Any one did something similar?

thank you for your advice and help.
ASKER CERTIFIED SOLUTION
Avatar of Wayne Herbert
Wayne Herbert
Flag of United States of America image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of yodaa
yodaa

ASKER

Thanks Wayne you are a star !! ]


Any another suggestions ?

thanks
Avatar of btan
Will also depends on RTO and RPO you have for the recovery time and backup point. Some instance include two sites as per planned for primary and DR site are identified in two locality apart and they are load balanced via DNS checks (e.g. using F5 GTM) for service and appls failover. Generally replication of application and data from the primary to the DR is over WAN. At time there may be separate vault facility as a backup for the primary and DR. This replication will initially consist of entire replication of applications and data, and follows a schedule replication of files/data.

Not to neglect to also develop redundancies to avoid single point of failures and factor in your plan: Hardware clustering and redundancy; Fail-over protocol or load balancing; Redundant network links; and ISP exchange diversity.

We also have another set of machine in the DR for the BCP and data are replicated across so there is no remote per se as we are planning for the worst case the primary is totally down. Do
consider using standards as part of the process, including NIST SP 800-34, ISO/IEC 24762, and BS 25777, as they provide a useful structured format for plans as well as guidance on the issues to address; this aspect is particularly important if plans will be audited.
http://searchdisasterrecovery.techtarget.com/Data-center-disaster-recovery-plan-template-and-guide
As btan noted failover processing is a necessity.  We have done this with AD, DNS, DHCP, and remote access services so that either in house server can fail and and everything keeps running.
 
Both servers are connected to switches with dual NIC's.  We use separate battery backup for each server so that if a backup were to fail we are still alive.  Each UPS is plugged into a different circuit as well.
 
We opted to go with virtualization but not clustering, for a couple of reasons.

  1. Clustering adds additional complexity, which is not justified for our small shop's needs.
  2. Clustering requires external storage in the form of a NAS, and this adds additional complexities as well costs.
  3. We can get the necessary data redundancy through DFSR, a well proven technology that also supports data deduplication.

We are contemplating going to dual WAN's, although cost is a consideration.  We've got Comcast cable right now and AT&T has just pulled new 100% fiber into the building.  Equivalent service for about $25 less per month than Comcast.  But, AT&T's DSL in the building uses the same fiber so there is no redundancy there.  We will either go with dual Comcast/AT&T, split between data and VOIP, with automatic failover in case one goes down, or we will choose a wireless LTE Verizon backup... lower data rates but cheaper and good enough to keep essential communications alive.

And as I have mentioned VOIP, even though we probably have enough staff to justify an inhouse PBX, we'll probably stay cloud, mainly because of the extra protection against building equipment failure, and the ability to readily use cell phones with office phone numbers.
Indeed redundancies are a major part of our plan and dual ring WAN with pairs of router are catered and maintained. We are into virtualisation too as most system are already segregated in Web/app and DB tier.   software clustering done at servers in each tier and load balancer appliances cater for hardware failover. Dual power (including UPS for full loading takeover) for each appliance and onsite standby unit catered to be exercised as immediate replacement

 We also have to make sure the planning norm factor in the security appliances and secure backup with point to point replication encryption via vpn for site to site  transfer. Not easily to strike for good performance but we are fine to go for optimal health as long as data are protected and can stay in compliance with audits.

If we fail to plan well with checks on the KPIs, we plan to fail on verification in event of real happening against what are assumed but not turn out as expected. So do conduct the exercises and test out woth involvement of system owner primary amd secondary contact.
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of yodaa

ASKER

Cheers guys