5

I am currently evaluating if I can use Amazon CloudSearch for our search need instead of Elastic Search. Right now, I just have about 4K small documents for my testing purpose. Whenever I ask it to reindex (lets' say after changing the instance type to m3.2xlarge from default), it takes a long time, a good 50-60 minutes to go to active status from processing status. Is it an expected behavior? Any suggestions on improving this?

vkumars
  • 163
  • 1
  • 6
  • I was going to say increase the instance type and numbers but testing on an m1.small instance with only 8 searchable documents, looks like re-indexing speed is not at all related to how heavy the load is, looks like there is a sort of time-out or maybe a queue to reach :/ Will continue looking into this – mim Aug 27 '20 at 22:01
  • 4
    I'm seeing similarly terrible times. The AWS docs claim that how long it will take to reindex depends on how many documents you have, and it could take from "a few minutes" to "several hours, if you have millions of documents". However I've seen it take 30+ minutes to reindex ***even when there's only one document*** – stevendesu Sep 02 '20 at 15:18
  • 2
    Same experience here. – Brian Cannard Jul 06 '22 at 15:45
  • 30+ minutes to reindex a single document (I personally can say this is true, having come here from google facing the same situation). I can't believe this was ever considered usable, there's something wrong here. – N.J.Dawson Feb 16 '23 at 14:57

0 Answers0