Datadog vs Firebase Crashlytics

Datadog is a no-brainer for any environment that relies on multiple servers. Even for one server, it's so much better to use for monitoring and alert capability than anything else I remember looking at before we started using Datadog. Teams can even use it to track their progress of investigating an alert so multiple people aren't unknowingly looking into the same exact issue unnecessarily.

Crashlytics is almost a defacto standard in the Android world for tracking crashes in the field. I've used it on the last 4 Android projects I've been on. It's not as helpful on iOS' stacktraces but I'm not sure if anything would be better because iOS stack traces are from ObjC code which is not as useful as Java stacktraces.

• APIs, the ability to interact with the data we pull into data dog is key. We port the information over to Servicenow, so the ability to pull everything into DataDog, then Servicenow, is a key component of our success here at Wayfair.
• Simple Interface - clean, useful, effective. Allows users to use DataDog for one reason, get work done.
• Lightweight agent on hosts

• Can get the device detail an actual state like RAM, Storage, Display Size, etc.
• Can get the class name and line number where crash occur.
• Its work in realtime so we can reduce the time of issue finding and resolution.

• User interface could be improved in some areas, looking for host information and the agent install link takes a bit of time.
• Configuration of the agent is generally done via a config file which is a pro and a con. It would be nice to have some UI to configure various agent options.
• Billing isn't entirely straightforward, they could use more reports to figure out the source of your monthly costs.

• Sometimes UI of Firebase Crashlytics is quit difficult to understand for new users.
• There can be a proper discretion for all the features that is offered.
• There can be api crash report also.

Because its fulfilling our basic and necessary requirements that can be an only and enough reason to renew the subscription.

The user interface is quite intuitive with the exception of the network map. As a deployer of software, it is trivial to setup.

The support team usually gets it right. We did have a rather complicate issue setting up monitoring on a domain controller. However, they are usually responsive and helpful over chat. The downside would be I don’t think they have any phone support. If that is important to you this might not be a good fit.

We are still trying other products, but people still like Datadog. After setting up a dashboard, it's great for monitoring instances on Datadog. Also, the DevOps team had a good time setting up Datadog. It means Datadog was way easier to set up compared to those others.

As mentioned previously, the only realistic competitor to Crashlytics is Instabug (backed by Y Combinator in the W16 class). While Crashlytics is focused on actual crashes, Instabug's main innovation is in collecting feedback from users for non-crash bugs. While non-crash bugs are a substantial category of bugs, Instabug's pricing is quite steep. As soon as you go beyond the most basic features your monthly bill spikes to $124 - $499 per month, whereas Crashlytics is 100% free (for all usage levels and all enterprise features). Based on this, I've never tried Instabug because the ROI just wasn't there.

• One of our critical production databases went down a couple of weeks back. Just because we had the Datadog monitoring enabled for that database, Datadog sent the alerts to PagerDuty and we were instantly notified via pager and email alert. We could act right away and bring up the database within a few mins.
• Ability to create custom screenboards have allowed us to track the critical and warning alerts in front of our eyes all the time

• Allows us to easily share crash data with clients.
• Provides confidence to our developers and project teams by having trust in the data that is being reported.
• Allows us to quickly and efficiently address fatal crashes.