Everyone ends up rolling their own Elasticsearch cluster to get as close to parity as possible at a lower cost.
Everyone ends up rolling their own Elasticsearch cluster to get as close to parity as possible at a lower cost.
FWIW, when I worked at BigCo Splunk was really great for standardization of logs and debugging across teams. We used it a lot when there were major outages to share unified ways across teams of slicing data and deep diving into logs together. I think it’s a fabulous product. The fact that they haven’t been able to monetize it effectively is a bit baffling since it doesn’t seem radically different than other B2B software solutions aimed at large corps.
The in-house elastic search clusters ended up being expensive to run and a burden with staff / hosting which resulted in rent vs buy and Splunk/Datdog being used to replace the in-house log storage.
In those situations both solutions ended running in parallel with some teams using one service over the other depending what was available when the services where developed and how much hassle it is to migrate. Log migration always being lowest priority for a team if it’s currently working and a backlog full of other stuff.