Our team is in charge of the network routing and quality of access of the different applications. As part of our mission, we need to visualize the flows towards our datacenter and to monitor the response times of all applications for each of the 300 remote sites.
In 2011, we started to look for a tool capable of addressing our needs. We selected a few products and tested them intensively on our datacenter: some proved to be good at protocol analysis but way too complex to manage for our needs; they would have required more dedicated resources than we could afford. All reports had to be created from zero and were not available “out of the box” (reporting was indeed an important feature for the Police Headquarters). Other solutions only covered a portion of our needs, because they were based on active QoS /compression devices and were not rendering the necessary diagnostic and analysis capabilities.
In the end we selected Performance Vision based on its strengths which fit well with our needs: In no more than a half day, we were getting relevant information and could use the product; The drill down access to information was fully intuitive; The tool was adapted to all our team (including punctual / light users); The network context (geographical and IP addressing plans) was perfectly translated into the interface and reports.
Moreover, Performance Vision offered an easy way to automate the distribution of graphical response time reports to all our IT staff located on each remote site; the information provided was easily understood by them and required little additional information from our team.
For diagnostics, Performance Vision outperforms our previous sniffer solutions and Netflow collectors! We started by making an assessment of all applications to understand their respective behavior: we pointed out a certain number of conception defects, which had a negative impact on users despite good server and network performance. We also built a baseline and determined the normal behavior of each application. We also implemented a daily reporting on our most critical applications to check that their performance andbehavior remained in line with the baseline.
Secondly, we proceeded to a large "house cleaning" operation: we got rid of large volumes of DNS errors, corrected the configuration of many workstations (which were using bad URLs or IP addresses for updates), as well as the one of certain servers relying on IP addresses instead of domain names, etc…
Performance Vision is an easy-to-use product; its administration requires very little time. Of course, Performance Vision provides data which has to be interpreted (e.g. one of our critical applications has very good response time per transaction but it involves a large number of transactions per user). The training is a key element in the project: as far as we are concerned, it got delivered in our context, on our own data. It has been very efficient and has quickly put us in a position where we can efficiently understand the information we are getting.
Overall, our experience has been brilliant: we consider Performance Vision as a very good product which fits our initial wishes. It is simple to use, provides a lot of relevant information. Upfront, it is very easy to use for network administrators and provides well designed reports for other IT staff members. We keep on enriching its configuration and using it on a daily basis. We have the project to extend its coverage through an additional virtual probe to analyze network traffic on remote sites (to analyze the application flows which cannot be viewed from our datacenter and diagnose local issues on remote sites)”.