DISASTER RECOVERY

"

Background As we all know, a potential business disorder will directly impact CAPEX and OPEX of the business. When a proper disaster recovery plan is set, any business can continue to run its critical operations without thwarting their customer expectations. That is exactly what we offered Indian defense, a solution that can manage and provide necessary outcomes in case of any DR issues.

Sector - Defence

Automobiles

BUSINESS OBJECTIVE

Transformation towards becoming hardware / Os / Vender / Operation agnostic.
Make everything software defined.

Technologies

AWS, Infrastructure as a code,
microservices, devops

Overview

As we all know, a potential business disorder will directly impact CAPEX and OPEX of the business. When a proper disaster recovery plan is set, any business can continue to run its critical operations without thwarting their customer expectations. That is exactly what we offered Indian defense, a solution that can manage and provide necessary outcomes in case of any DR issues.

Challenges

  • The Indian defense had only 4 Mbps site to site leased line for data transfer.
  • They were using email, intranet, and few BPM applications on SAN-based storage.
  • They were not comfortable to use array-based replication neither the FCIP based switch communication.
  • They were expecting a lightweight, script-based DR solution, that is managed with ease and extended as per requirements and with the SLA that is less than 10min.

Solutions

  • Thorough planning and execution proved to be successful. Firstly, we used copy-on-write technology advantage that takes an incremental snapshot on a purpose-built storage.
  • We performed deduplication of that snapshot and transferred across the site by simple Zfs send-receive.
  • In case of failover, we could recover application instantly- since it is mare mount of the filesystem, so recovery time was almost zero.
  • We used some open source tools that work together with some simple scripting and automation and deliver the required solution.
  • The solution is extracted from components like snapshot engine- pool provisioning –Zfs based storage and virtualization technique.
  • The all in one solution comes with the freedom to unlimited extension/customization.

That old Cliché “DR is resilient enough to implement in one go” we proved it wrong! Designing and deploying a total DR solution was not a rocket science for us. However, there were few challenges to face though; we had only 4 Mbps site to site leased line for data transfer. Due to pricing aspect, the Indian defense was neither ready to accept array-based replication or the FCIP based switch communication. They wanted a lightweight solution and a script based DR solution to manage their tasks effortlessly. With all these hurdles on the path, we managed to script a DR solution that is open and restores the data in DR site. In case of a network latency or a failure, Indian defense could recover the application in no time and the recovery time is almost zero.

OUTCOME

  • The DR solution was 10% low compared to a licensed DR product.
  • They were able to restore the data instantly and the recovery time was less than 10 min which is almost zero.
  • The lightweight script was easily adaptable and manageable.