3 ms·
1. You have to keep your logging, monitoring, alerting infrastructure up yourself 2. You have to monitor, log, and alert on your that infrastructure yourself w
by jjjensen90 7y ago
1. You have to keep your logging, monitoring, alerting infrastructure up yourself
2. You have to monitor, log, and alert on your that infrastructure yourself with something else
3. You usually have to spend more money, both in hosting and in developer/ops time, on getting something mediocre compared to a provider that exclusively does observability as a product
4. Logging etc are a commodity and you should have a really good reason to build or run something yourself if it is not your core competency
5. Observability is much harder than it sounds and providing a cohesive platform that ties log events to traces to metrics to alerts, and keeping it up and highly available is a hard problem that you shouldn't do if observability isn't your core competency/product
- tedk-42 7y agoYep, if you can afford it make it someone else's problem. No team should need to have dedicated OPs people that are responsible for keeping the logging platform running. This is a solved problem and costs the company more in engineering time (paying a person to do the job) than the Saas solution. E.g. 30K/month isn't much when you compare it to paying a sole engineer 15K/year + cloud costs of hosting your own logging solution + any licenses. There's also a slim chance the home grown logging platform will be as resilient as a Saas product. I for one am very grateful to DataDog's support team as they've been very helpful in debugging logging problems we've had in the past.