Hacker News Reader

Comment by abiro | Hacker News Reader

Don’t read too much into this — they are just catching up with GCP and Azure here. AWS has a history of introducing half-assed products with great fanfare when the lack of such a product could be a disadvantage for customer acquisition (think Beanstalk, Cognito or Amplify).

View on HN

Don’t forget their ElasticSearch version. Now that is a shit show.

what do you mean? they have customers using it in prod

So did MongoDB from day minus one.

Well mongodb is a much better product than it was a few years ago. not saying its a good one but...

and aws elastic is just a fork of a stable elastic release. it makes sense if the cost is competitive

There are... quite a number of reasons you should give pause to the idea of running AWS hosted elastic.

If you’re really up against the wall and don’t want to manage elasticsearch I would highly recommend getting the managed services from Elastic themselves. Support is decent and they’re cost competitive.

> There are... quite a number of reasons you should give pause to the idea of running AWS hosted elastic.

Care to share some of these reasons?

I can't go into details, but let me say that someone who would be in the know at AWS, very much off the record, told us not to use their hosted elastic because of how terrible it is.

Some issues: scaling it is manual (a "managed service" that requires direct management to scale is not a managed service lol), the elastic versions supported are usually woefully out of date, limited customer adoption means that the product doesn't get a lot of engineering attention and is just limping along, etc.

That’s what we were told as well. Plus also various things like cross region replication being unavailable. Also things like management ports are not possible to get to.

Edit: also plan to lose everything in it.

I totally agree but if your requirement is just to have elastic for a standard search use case and don't need any new features it might make sense... out of date doesnt mean its garbage.

I am not against manual scaling anyway at some level you always need to do manual scaling on AWS whatever the product...

Their sales team are awful. They don’t even understand the product they are selling.

AWS could release a literal turd and someone somewhere would use it in production. Which doesn't mean you should too.

Man I remember, whenever we needed to update a cloudformation template that triggered an ElasticSearch modification it was horrible. And of course, after waiting 1.5h, some dns change failed and it had to be rolled back and you just wasted half a day.

Fun times.

That reminds me of that time I had to do a snapshot on an EBS volume that had Jenkins on it. Took five days to do a 2TB volume.

Local VMware about ten minutes.

Edit: I notice a flurry of downvotes arriving since US woke up. Must be Amazon team coming on line :)

AWS mislabeled something that is a "backup" (delta copy to s3) as a "snapshot". That's not what is called snapshot in the storage world. Granted ESX snapshots aren't exactly that either.

Why is it a shit show?

This article neatly outlines some of it. Unfortunately I didn’t go digging and got enticed by the marketing in the end.

https://spun.io/2019/10/10/aws-elasticsearch-a-fundamentally...

I can't think of a better description of their EKS launch than "half-assed product introduced with great fanfare when the lack of an alternative to GKE was a disadvantage for customer acquisition." EKS had much improved when I last played with it around a year ago, though.

Reply on news.ycombinator.com