Millions of patients and healthcare professionals worldwide depend on Ambu’s products, often in life-threatening situations. Ambu’s products are in such high demand that manufacturing and warehouse facilities in China, Malaysia, the United Kingdom and the United States operate around the clock. Their operations depend on the availability of Ambu’s enterprise-resource management (ERP) system, which is part of a massive IT infrastructure spread across 10 data centres worldwide. Ninety-nine percent of the IT infrastructure is virtualised on VMware vSphere. Approximately 400 virtual machines (VMs) contain 50TB of data.
“Ambu never shuts down,” said Bo Norman Wilhelmsen, Network Administrator at Ambu A/S. “If our ERP system isn’t available to our 1,000 users at all times, production and shipping will stop. Downtime could cost approximately $30,000 per hour.”
The VMs running the ERP system are in Ambu’s primary data centre located in Denmark. If a disaster strikes the primary data centre, critical systems including the ERP system must be restored quickly in the secondary data centre to resume business operations. Ambu’s disaster recovery (DR) strategy relies on reliable backup and replication combined with high-speed recovery.
“Our backup tool didn’t work as promised,” Wilhelmsen said. “We never got replication up and running, and backups failed often. One time I tried to restore email from a backup of Lotus Notes. The backup was corrupted, yet the backup tool reported the backup was successful.”
Too many failed and unrecoverable backups made it nearly impossible for Ambu to meet internal service level agreements (SLAs). Another challenge was backing up all VMs within the backup window. Backup wasn’t slow, but maintenance was slow. The backup tool compressed and deduplicated data after backups completed rather than on the fly; therefore, backups could not run during maintenance. With data doubling every two years, Ambu knew it would be difficult to back up even more VMs within its backup window.
“I contacted the backup vendor for help in speeding up backups, and I was told to purchase faster hardware,” Wilhelmsen said. That was the breaking point. I was spending 80% of my time troubleshooting backups when I should have been focusing on other IT projects.”
Veeam® Backup & Replication™ supports Ambu’s DR strategy with fast and reliable backup, replication and recovery. Because Veeam automatically verifies the recoverability of every backup and replica, Ambu knows all virtualised data and applications, including the ERP system that runs production and warehouse facilities, are available 24x7.
“Veeam gives me confidence that our virtualised systems are available so our business remains operational at all times,” Wilhelmsen said. “I wasn’t confident I could recover even one system with the previous backup tool.”
Wilhelmsen said the previous backup vendor was wrong about Ambu needing faster hardware for faster backup.
“Faster hardware was definitely not necessary—Veeam works great on our existing hardware,” he said.
Ambu uses Veeam in 10 data centres across the globe for onsite backup of 400 VMs containing 50TB of data. Veeam replicates VMs to the primary data centre in Denmark for fast failover and failback in case of disaster. Veeam’s Built-in WAN Acceleration optimises replica transfer because some data centres have limited bandwidth, making it unnecessary for Ambu to deploy a general-purpose WAN acceleration appliance or acquire additional network bandwidth.
Ambu easily meets SLAs with Veeam. Reliable replication provides multiple, verified restore points and high-speed recovery restores whole VMs, files and application items in minutes. Veeam also enables Ambu to back up all VMs within its backup window because Veeam deduplicates and compresses backup data on the fly rather than waiting until all backups have completed like the previous backup tool.
Wilhelmsen said he and his colleagues were grateful when Veeam recovery prevented a virus from spreading.
“We had replicated 400GB between data centres in Denmark and China, and half that data became infected by a virus in China,” Wilhelmsen explained. “We used Veeam to quickly restore files based on file type from a recent backup and avoided spreading a virus that could have impacted business operations.”
Proactive issue resolution is key for 24x7x365 availability; therefore, Ambu deployed an additional Veeam solution—Veeam Management Pack™ (MP) for System Center. Veeam MP integrates seamlessly with Microsoft System Center to provide complete visibility of Ambu’s physical, virtual and backup infrastructures from the System Center console.
“Being able to monitor all environments under a single pane of glass greatly reduced the time needed for pinpointing issues,” Wilhelmsen said. “Before Veeam MP, we used three different monitoring tools, so it took a while to figure out if what we were seeing with one tool was the same with another tool.”
Veeam MP helped Ambu identify and resolve issues in two situations that could have threatened business operations.
“Sometimes we have to debug the virtual environment and SAN layers, which isn’t optimal for a 24x7 business. With Veeam MP monitoring, we can see relationships among resources so we can locate and fix problems fast to maintain 24x7 availability.”
Another time, Veeam MP helped Ambu detect a VM that overused storage input/output (I/O).
“High I/O affected additional VMs, including the ones that run our ERP system, which utilises SQL Server. Microsoft System Center Management Pack for SQL Server helped us see a bad memory configuration in SQL Server, so we resolved both issues fast. That single pane of glass is good for us.”
In addition to providing real-time monitoring, Veeam MP gathers and correlates data into several reports that help Ambu, such as capacity planning reports for planning resources in the virtual environment (especially for the ERP system) and reports that outline SLAs that have been met per system.
“We share the SLA reports with management so they know the business is safe,” Wilhelmsen said. “Using Veeam MP and Veeam Backup & Replication give us confidence in our IT infrastructure and saves time. With Veeam I spend 10% of my time on backup and recovery, which is down from 80%.”
Supports DR strategy with 24x7 availability
If Ambu’s primary data centre suffers a disaster, the IT team will use Veeam Backup & Replication to recover a corrupted VM by failing over to its replica at a secondary data centre. It multiple VMs fail, Ambu will use batch processing to restore operations with minimum downtime.
Provides automatic recovery verification and fast restore
Ambu automatically verifies the recoverability of backups and replicas with Veeam’s SureBackup® and SureReplica. Veeam starts a backup or a replica in an isolated environment called a Virtual Lab, performs a series of tests and creates a report about its state and condition. Fast restore features such as Instant VM Recovery™ let Ambu restart a failed VM from the most recent backup. Veeam Explorers™ for Microsoft Active Directory and SQL Server restore application items.
Facilitates single pane of glass for proactive issue resolution
Veeam MP extends System Center for application-to-metal visibility, giving Ambu the complete picture of its entire IT infrastructure. With Veeam MP, Ambu identifies and resolves issues before they become problems and impact the business.