What does your network have in common with a cloud service provider? At first glance, you would probably say not much. But remove the word cloud, and you have an over-taxed network team spending several hours troubleshooting complex database, Citrix, VoIP, and Internet performance problems over a highly-distributed network. Too many problems and applications and too little time to manage them are challenges that will sound familiar to any network professional.
Additionally, because it’s a cloud provider, they have client service level agreements (SLAs), which raise the stakes for the network team to find and fix problems quickly, or face paying money to their clients for downtime. In this high-tension environment, German telecom and cloud service provider GELSEN-NET, was able to turn their monitoring strategy around to achieve an 87% reduction in troubleshooting time. Through this shift to rapid resolution, we’ll explore:
- Critical challenges faced when managing performance with freeware
- Criteria for choosing the correct RNA solution
- ROI and time-savings benefits
Founded in 1983, GELSEN-NET has grown by providing comprehensive Internet connectivity, phone, and IT services to over 72,000 residential and business customers. Over the past two years, GELSEN-NET has added extensive cloud services to its IT offerings, providing local businesses with external hosting, access to scalable storage, and cloud-based disaster recovery.
The shift to the cloud has brought about significant new challenges for Bjoern Ferber, the network engineer in charge of managing performance, ensuring quality of service, and validating SLAs with clients. To ensure on-time execution of cloud-hosted applications and meeting SLAs, not only did the company need to be sure their network was performing at premium levels, they required flexibility and visibility into service performance from their datacenters to clients' sites.
This became clear when two of GELSEN-NET's larger cloud customers hosting LDAP and SQL database traffic in the carrier's datacenter began experiencing slower response times between the hosted servers and external clients. Ferber and his colleagues attempted to identify the cause of slow response times with Wireshark network analyzers, but lacked the application-layer visibility to find the issue between the servers, network, and clients.
"Wireshark provides a lot of information, but isolating problems requires significant upfront knowledge that we simply didn't have," said Ferber. "You need to know the MAC and IP addresses along with packet payload details. Given the number of clients we have and the very nature of data being warehoused in the cloud, there's a lot to sift through. Navigating through large trace files was an unmanageable, waste of time."
Ferber knew GELSEN-NET needed a flexible solution that could provide visibility and monitoring intelligence to resolve performance problems faster. Network consultants GORDION Data Systems Technology answered the carrier's call with Network Instruments GigaStor retrospective network analysis appliance, which captures terabytes of traffic and applies advanced analytics to resolve performance challenges. They chose the portable GigaStor appliance because flexibility was critical, allowing Ferber's team to begin troubleshooting problems at their datacenters and go offsite to the client when necessary.
To select a monitoring solution, the network team created test scenarios around a common Citrix connection issue. In running the scenarios, GELSEN-NET's network team knew it was either a bandwidth issue or there were too many Citrix connections on the server. With GigaStor at the customer's site in the test situation, they were able to capture traffic, generate a filter, isolate the issue to the server, and provide resolution in an hour.
"Compared to other monitoring solutions, GigaStor allowed us to resolve problems faster," said Ferber. "It was very intuitive and had a lot of features that quickly pinpoint problem causes. Competing tools weren't as user friendly to manage or in utilizing various analysis features."
Rather than running around for several days with two laptops and copies of Wireshark to solve problems, Ferber's team is now able to track and resolve the same issues in less than 60 minutes using GigaStor. This has yielded a time savings of up to 87 percent, spending only an hour today on root-cause analysis where 8 hours were spent in the past.
Solving performance challenges faster also protects GELSEN-NET's bottom line. With strict SLAs in place, Ferber's team has between 5 to 8 hours to resolve the performance problem, otherwise the service provider risks paying the client money for downtime and outages. "With fast problem solving, our GigaStor has quickly paid for itself," said Ferber.
Learn how the company uses GigaStor Portable to manage VoIP rollouts and troubleshooting communications traffic. To understand how RNA can slash troubleshooting time, check out the following video.
Thanks to Network Instruments for the article.