“Coop is expanding and is depending on fast growth”, says Michel Rodel, Head of the Unix Solutions Infrastructure Team, Coop. The company had therefore standardised and modernised their vSphere-based virtualisation infrastructure to strengthen IT and reduce complexity.Coop has two data centres, one each in Basel and Bern and had been looking for a long time for a disaster recovery solution that could be implemented more flexibly than the existing environment that was based on VMware's SRM Manager.
“SRM didn’t give us the flexibility we needed. The distance between the two sites meant we had to choose an active/passive configuration and protect them with an asynchronous storage replication with a round trip time of 1.5 milliseconds – this was causing noticeable issues”, adds Rodel. This was particularly the case with regard to Coop’s All-Flash storage infrastructure that achieves times of under a millisecond. In addition, the existing solution only protected the most important systems.
The object was to complete secure both data centres with the disaster recovery plan with tape backups created every seven days providing extra security in the future. The time in between would be covered by Zerto, which is why Coop renewed the backup software in 2017 and setup a new tape library.
Zerto Virtual Replication was key. The solution minimised complexity and has made our IT environment more flexible.
Michel Rodel, Head of Coop Unix Solutions
Process. Coop first approached Bechtle in 2016 to plan the future of their disaster recovery strategy and evaluate a new DR solution Together, Coop and Bechtle put together a Statement of Works and a call for tenders which initially attracted four candidates. The tendering process brought two solutions to the fore: At the end of 2017, a proof of concept (PoC)was set up for both Zerto Virtual Replication and another solution and tested against the detailed catalogue of requirements.
“The Zerto PoC was set up within half a day”, recalls Dario Dörflinger, Senior Systems Engineer, Bechtle. The Zerto environment was configured the following day and then data synchronisation could begin and in-depth testing carried out. The solution fulfilled all the requirements very well and, after the testing process, Coop decided to go with Zerto.
Solution. Zerto Virtual Replication enables the automation of implementation and management as well as the testing of the disaster recovery plan for virtualised environments. Zerto’s own technology shifts replication and recovery to the hypervisor level, which, according to the vendor, is much more flexible, affordable and easier to control than a storage and LUN-based DR solution.
Coop’s experience shows that the promises were kept: “Zerto was key. The solution minimised complexity and has made our IT environment more flexible. Decisive factors in the decision were the easy installation, simple operation, product quality and scalability”, sums up Michel Rodel.
Thanks to the new solution, Coop can easily move workloads 139 kilometres from one data centre to another. Zerto reduces complexity in a number of ways: Replications are now on the datastore virtual disk level instead of LUN-level: “This means that we can replicate as dynamically as we like”, says Rodel. A variety of services can be consolidated in a Virtual Replication Group (VPG) which makes management much easier.
Failovers used to occur over the entire cluster, but now are more flexible and less complex on the VPG-level. “In short. Our IT department can focus more on design and architecture and doesn’t need to worry so much about services and support”
We were also very impressed by the vendor’s support. Even during the proof of concept, Zerto sent a specialist to support us and when we had support low priority support requests, they were responded to within minutes.
First and foremost, Zerto Virtual Replication is already protecting applications at the Coop headquarters and this will be expanded to the vHANA environment in the future. 2,800 machines are currently covered.
Zerto’s Continuous Data Protection feature creates various checkpoints that don’t negatively affect performance. This guarantees minimum data loss in the event of errors such as data corruption or incorrect updates and upgrades. The first figures are impressive: With an RPO of only a few seconds in the first 15 minutes and an RTO of 2.31 minutes, Zerto’ is providing Coop with a number of benefits compared to the previous solution. In the past, backups were run every 24 hours, now it’s every 15 minutes.Michel Rodel has one wish: “It would be great if we could also use Zerto for long-term backups. We are currently utilising another solution as we are relying on tape support.” Due to cost reasons, tapes remain a strategic solution for Coop.