Perform advanced search in your application logs

News
1

Hi,

Having to perform full-text searches on old logs, storing them for several years, applying complex filters, and so on has been requested several times.

This is why we’ve built tutorials for those use cases and even with Grafana! Here are the tutorials:

  1. Integrates Cloudwatch with Qovery to store and read your application logs
  2. Use Grafana with Cloudwatch
  3. Use Datadog to store and read your logs

More will come directly inside the product in the coming weeks so stay tuned!

3 Likes
2

Hi @Pierre_Mavro thanks for your tuto. I have a question, I followed step with fluent-bit and some pods can’t start because of insufficient resources but I got this error pod didn't trigger scale-up: 1 node(s) didn't match Pod's node affinity/selector what can I do to enable this namespace to trigger scale up ?

Thanks

3

Hi,

Sorry for the late answer, I guess your cluster nodes do not have enough resources. What happens if you do a describe on the daemonset? Are the pod in Pending state?

Thanks

4

Hi, I don’t know, I just follow the guide you wrote and some pods can’t start due to this error. At this time it’s fine because a service manage by Qovery trigger a scale up on a new node so there is anough ressources.
But maybe there is a flag to set on fluent-bit installation to trigger an autoscalling by Qovery ?

5

As it’s deamonset, if one node does not have resources, you won’t have all the fluent-bit pods appropriately deployed. Horizontal scaling will not help in this case :frowning: .

In this case, you need to have bigger nodes or applications rescheduled (ex: changing instance type to force applications re-allocation) on the whole cluster to let fluent-bit pods boot properly.