[dirtshell]

Do any engineers actually like DD? Execs/managers seem to love DD and get upsold all the time on it and ask engineers to implement some of their half-baked features. It seems like its good for alerts and dashboards for infrastructure teams. But as an engineer its a pain and using it for log analysis is much more annoying than just tracking down the actual relevant logs and grepping.

[everfrustrated]

Easily the best product in its category. The value is in APM/tracing - if you're just using it for logs you're missing a lot.

If you're used to traditional Enterprise pricing it's fairly priced for the value you get but anybody coming from self-funded or VC it's very expensive. If you're already using Splunk you can afford it.

One of its best features is it's consumption priced not by seats so easy to open up for all of eng including product/QA teams and not just devs to use which is great for breaking down barriers.

It's a bit like aws in that it's a platform - use of one product tends to encourage using more from their suite.

[xtracto]

Second this.

I used to log large apps using kibana and elastic search. Also using the clusterfuck that are all AWS tries at this (cloud watch, log insights.amd whatnot).

Nothing compares to what DD give you in observability.

Having said that, DD should only be an AWS feature. They should buy them for a couple of billions and integrate it as a service for all of AWS infrastructure.

[leeoniya]

> They should buy them for a couple of billions

DD has a market cap of $48B

[hobs]

Which just meant that when the business had sticker shock they just disabled all the useful features of DD which consumed so much.

[jitl]

It’s good, I like it. I don’t use the logs product (we splunk) but I’m big fan of the automatic profiling and trace/span stuff. It’s like 50% better UX compared to other tools I’ve tried. But it’s expensive enough we’re always thinking about moving off it. That will be a very sad day for me.

[SteveNuts]

My biggest issue with them is the absolutely, hilariously, ridiculously expensive custom metrics.

[geraldwhen]

Large json logs just won’t appear for hours, which makes it so fun to debug issues in deployed environments!

Showing ALL the logs isn’t cool. Showing SOME logs at random is cool.

[hackernoops]

DD is the Oracle/SAP of o11y. Scumbags that kill innovation.

[datavirtue]

It's just a solution to garbage distributed cloud service logs and devs that don't know how to log an application.