-
I am using Dexie in my Firefox add-on, Clippings. I'm also using Dexie.Observable, but when I recently submitted a new version of my Firefox add-on, it was rejected because Dexie.Observable is in beta…
-
### Describe the question/issue
Experiencing seemingly a fluent-bit related bug (low frequency and sporadic) where in the FB pod is not correctly sending logs from the node. Additionally the nod…
-
### Context
Kubernetes is one of the richest systems in the number of fields available for observability making it susceptible to inefficiencies given the potential for high cardinality of components…
-
**Describe the bug**
When IDMSv2 is enabled in the worker nodes with hop limit 1 , the IMDSv2 is not accessible from the pods. In general, I don't want pods to access IMDS since they can get credenti…
-
### Describe the question/issue
I want to customize the Fluent Bit image used in the EKS on Fargate setup.
I set up logging to S3 via Kinesis Data Firehose as documented here: https://docs.aws.a…
-
Add refresh functionality to all Observability alert detail pages.
## Use cases
- When alert is recovered, but the status is not updated on page.
- When some time has passed since user opened a…
-
-
### Describe the bug
OpenTelemetry instrumentation is still active even when the SDK is disabled, therefore performing a non trivial amount of work and allocations that are not needed.
See the f…
-
**Goals**
In order for CAPI to work it needs credentials to talk to the infra providers, which have quite the high amount of permissions. We need to ensure that such credentials can only be seen by a…
-
We noticed that lm_eval --model vllm did not work when data_parallel_size > 1 and got `Error: No available node types can fulfill resource request` from Ray. After some research, I believe when `tenso…