We have 45 million documents and i am wondering how can we index all of it? we use version 5.5
Welcome to the forums and nice work on snagging a cool nickname.
The answer depends quite a bit on a few factors. I think first and foremost, it’s a function of what kind of system resources you have. Index building from no index will probably mean it’s somewhat dependent on disk IO, and our system intentionally prioritizes other operational requests over this kind of re-materialization. After that, it’s probably dependent on network IO depending on how you’ve laid out your cluster with the indexer nodes, and then after that it’s a function of index complexity (are you talking about primary index?) and CPU.
My colleagues @korrigan.clark and @wayne may know if we have anything that gives you a baseline with a specific configuration.
Finally, version 5.5 is really quite old at this stage. I would look into upgrading before adding indexes if possible.