In order to keep everything running smoothly, the company utilizes a series of additional Tableau tools and applications in conjunction with Tableau Server.
“We have a whole server ecosystem,” says Tim. “We have our disaster recovery server for standard failover events, plus our test server that we use primarily for testing upgrades before moving them to the production server.”
“We also have a utilities server that’s used for monitoring the production server and testing the test server. A variety of Tableau open source applications are installed there to help us with this, including TabMon, for production server monitoring, TabJolt, to stress test the test server when preparing for upgrades, and LogShark, to ingest log files from the production server.”
The whole ecosystem is run and monitored using a series of 45 Tableau dashboards, giving a complete overview of all hardware and applications, as well as allowing permission auditing and content organization. This allows Tim and his team to quickly pinpoint the causes of any issues as they arise, boosting productivity and avoiding unnecessary downtime.
“There have been multiple times where an individual user has pinged me to say the performance is pretty slow and I can quickly look at the dashboards and say, ‘it's just slow for you,’” says Tim. “Then, I can dig deeper and tell them it’s because the workbook they’re connecting to has a live custom SQL query against an underpowered database.”
“There’s a lot of ways we’re able to respond quickly because of the tools we've built with Tableau,” continues Tim. “These range from individual HTTP requests like the one I just mentioned, right through to when our data warehouse ETL process runs too long, impacting extract creation. When that happens, Tableau alerts us to the issue right away, so we can quickly get to the bottom of it.”