-
Hi, I’m having trouble installing `dune` using `opam`.
I’m trying to installing camomile with
```
opam install camomile
```
But as it depends on jbuilder, which in turn depends on dune, the…
-
Good day.
In one of the clusters, we had some problems with etcd, after restoring the work of etcd / api-server, the node-problem-detector failed to update its state and stuck in this state. Restart …
-
**Affected product area (please put an X in all that apply)**
[ ] Configuration Infrastructure
[ ] Docs
[ ] Installation
[X] Networking
[X] Performance and Scalability
[ ] Policies and Telemet…
-
The resource dashboards use inconsistent metrics for displaying memory usage. Currently the cluster dashboard uses RSS and the namespace one uses total usage. I would propose that both use working-set…
-
## Bug ##
It seems that multiple OOMkillers invoked in close succession cause ext4 filesystem to be remounted readonly due to journal corruption.
It was producing coredumps in parallel, so it mi…
-
## This is a...
[ ] Feature request
[ ] Regression (a behavior that used to work and stopped working in a new release)
[X] Bug report
[ ] Documentation issue or request
## Description
…
-
We provide various CLI tools, most of which are lightweight and perform relatively quick operations. By default, we do not set any java options when we launch the jvm to run these tools so they run wi…
-
**Description**
Enable integration with Stackdriver so it will be able to scrap Prometheus metrics from our long-running clusters (like nightly and weekly).
AC:
- Integration enabled
…
-
Hi,
Our pilot instances experienced massive spikes in memory usage immediately following the control plane upgrade and got killed by OOMkiller a few times:
This was presumably in the window t…
Stono updated
5 years ago
-
I have a pod consuming 115MB of RAM. I tried to set limits of 128MB, 256MB etc but the lowest one that worked was 450MB. Looks like the runner itself requires this amount of RAM, so the pod gets kille…