Conducting iSeries disaster recovery planning, regardless of the scale of the applications and data being handled, should start with a well-defined threat assessment matrix. This matrix includes physical infrastructure loss, communications downtime, power outages, and security breakdown.
Because the iSeries range of mainframes is found in all sorts of data centers, there is no single solution that can be applied to physical infrastructure loss of any scale. But in the worst-case scenario—that of destruction of the data center due to fire, flooding, or other acts of God—it goes without saying that a redundant site or secondary data center should be put in place in a remote location. Basic office facilities should be set up as well to accommodate those who will be working on disaster recovery.
If telecommunications lines go down, there must be alternative ways of gaining remote access to data. While having access to satellite communications is ideal, it is far too costly for small and medium-sized enterprises. Telecommunications providers should provide concrete plans for rerouting traffic from one point to another; if there are no clear plans, though, the secondary data center should have bandwidths and network speeds comparable to those in the primary site.
Ideally, a Tier-4 data center should have power coming from at least two utility providers. But in areas where power distribution is controlled by a single entity, it might be close to impossible. In this case, generator sets and a fuel supply should be on standby. The amount of backup power should be sufficient for the iSeries mainframe, application servers, storage, switches, and routers. There should also be enough power for the precision air-conditioning system to prevent the loss of machines due to overheating as well as the physical security devices and fire detection and suppression system.
Intruders might take advantage of power outages and other disasters by bypassing or breaking into physical security systems. The result is theft of both equipment and data. Thus, additional manpower or backup physical security systems should be deployed as soon as a disaster strikes and should be integral parts of any iSeries disaster recovery plan.
Once the threats above are assessed, it’s time to create a triage of applications and data that should be recovered first. This is because backup systems must accommodate large power and telecoms requirements at startup. Thus, a list of mission-critical applications should be drawn up with the help of machine owners and other stakeholders. In addition, organizations should make disaster recovery plans readily available and highly visible. They should also be ready to deploy additional personnel in case of a disaster.
Many organizations of different sizes choose to entrust their iSeries disaster recovery capabilities to an external provider instead of shelling out a large capex bill on backup data centers. An external provider, especially one with multiple sites, can provide recovery services regardless of the extent of damage sustained by a client’s primary data center. In addition, because personnel are already on site at the time of the disaster, the external provider can perform failover as soon as they get the go-signal from the client.