SuperQ
Repos
139
Followers
305
Following
8

Prometheus style smokeping

342
45

Recipes I've collected

17
2

Events

issue comment
Introduce a new opentelemetry plugin for observability

Also adding OpenTelemetry into CoreDNS allows even third party plugins to further built and extend observability capabilities of their plugins on top of OpenTelemetry.

This sounds a bit like a solution in search of a problem. What actual problems does this solve for CoreDNS?

For example, the existing Prometheusmetrics plugin allows all CoreDNS plugins to integrate into that observability framework.

Can other plugins not use the existing trace plugin?

Binding observability capabilities of CoreDNS only to one concrete technology (Zipkin) doesnt sound right to me.

But this is just tracing. Is tracing really so complicated that we can't just produce a format that other systems can consume it?

Again, I feel like this is a solution in search of a problem. Why is this a separate plugin and not changes to the existing trace plugin?

Created at 17 hours ago
issue comment
Add packet timeout option and make latency more accurate

FYI: This project is abandoned as we are unable to maintain it due to lack of maintainers with admin rights on the repo. See #221

Created at 1 day ago

Add ansible-lint to CI

Signed-off-by: SuperQ superq@gmail.com

Fix up ansible-lint issues.

Signed-off-by: SuperQ superq@gmail.com

fix(ci): lint action got moved from ansible-community

Signed-off-by: gardar gardar@users.noreply.github.com

refactor(ci): move collection dependencies to where ansible knows to pick it up

Signed-off-by: gardar gardar@users.noreply.github.com

fix(molecule): rename deprecated crypto modules

Signed-off-by: gardar gardar@users.noreply.github.com

fix(lint): fatal lint errors

Signed-off-by: gardar gardar@users.noreply.github.com

fix(lint): warning lint errors

Signed-off-by: gardar gardar@users.noreply.github.com

fix(lint): fix all linter exceptions and remove exception rules

Signed-off-by: gardar gardar@users.noreply.github.com

test(ci): check if collections path env var is needed

Signed-off-by: gardar gardar@users.noreply.github.com

fix(ci): colored output

Signed-off-by: gardar gardar@users.noreply.github.com

Merge pull request #6 from gardar/gardar/lint

Fix linting issues

Merge pull request #3 from prometheus-community/superq/lint

Add ansible-lint to CI

Created at 2 days ago
pull request closed
Add ansible-lint to CI

Signed-off-by: SuperQ superq@gmail.com

Created at 2 days ago
SuperQ delete branch superq/lint
Created at 2 days ago

fix(ci): lint action got moved from ansible-community

Signed-off-by: gardar gardar@users.noreply.github.com

refactor(ci): move collection dependencies to where ansible knows to pick it up

Signed-off-by: gardar gardar@users.noreply.github.com

fix(molecule): rename deprecated crypto modules

Signed-off-by: gardar gardar@users.noreply.github.com

fix(lint): fatal lint errors

Signed-off-by: gardar gardar@users.noreply.github.com

fix(lint): warning lint errors

Signed-off-by: gardar gardar@users.noreply.github.com

fix(lint): fix all linter exceptions and remove exception rules

Signed-off-by: gardar gardar@users.noreply.github.com

test(ci): check if collections path env var is needed

Signed-off-by: gardar gardar@users.noreply.github.com

fix(ci): colored output

Signed-off-by: gardar gardar@users.noreply.github.com

Merge pull request #6 from gardar/gardar/lint

Fix linting issues

Created at 2 days ago
pull request closed
Fix linting issues

Fixed the lint ci and finished fixing the linting issues in #3

Proposing this to be merged into #3 which can then be merged into master.

Created at 2 days ago
issue comment
Introduce a new opentelemetry plugin for observability

While it doesn't look like much, the imports bring in about 15k LoC in dependencies based on a scan of the changes to go mod vendor. I would be interested in how this compares to simplifying our trace plugin to a more dedicated tracing library like zipkin-go.

Created at 2 days ago
issue comment
Introduce a new opentelemetry plugin for observability

I'm not sure I want opentelemetry in CoreDNS. I'd rather go with a lighter weight tracing library. OTel brings in a lot of excess baggage.

Created at 2 days ago
Still maintained?

We have a new repo here: https://github.com/prometheus-community/ansible

Created at 2 days ago

Add cisco BGP example.

Signed-off-by: SuperQ superq@gmail.com

Created at 2 days ago
Speed up node rules

:eyeglasses: @paulfantom

Created at 3 days ago
Speed up node rules

Improve the speed of computing node:node_num_cpu:sum

  • Single pass count().
  • Pick one mode label, cuts cardinality selection by 8x.

Improve cluster CPU utilization computation

  • Add per-node utilization recording.
  • Simplify cluster computation by averaging over node pre-computing.

Signed-off-by: SuperQ superq@gmail.com

Created at 3 days ago
create branch
SuperQ create branch faster_cpu
Created at 3 days ago

Update node memory metrics

Use modern kernel "MemAvailable" metric when possible.

  • Update recording rule name.
  • Fallback to old MemFree+Buffers+Cached method for older kernels.
  • Also include Slab memory0.

Merge pull request #282 from SuperQ/superq/mem_available

Update node memory metrics

Colossal simplification of regexp's

More importantly, of course, this help to remind people that Prometheus regexp's are always anchored.

Signed-off-by: beorn7 beorn@grafana.com

Merge pull request #283 from beorn7/beorn7/regexp

Colossal simplification of regexp's

Add missing cluster filter to network panels

  • Fix Compute Resource dashboard networking panels to filter by $cluster, like for CPU and Memory panels. This is important when showMultiCluster=true bc the network panels are causing fanout to all clusters, erroneously aggregating across all of them, and may load a quite large number of pods into tables (slow)

Rel: https://github.com/kubernetes-monitoring/kubernetes-mixin/pull/248

Fix "Compute Resources / Workload" workload filter

  • Fix network panels in the "Compute Resources / Workload" dashboard to respect the selected "$workload" like CPU and memory panels

Fix typeTemplate to not hardcode datasource

  • Previously, "Compute Resources / Namespace (Workloads)" dashboard was missing a workload "$type" template variable, if there was no datasource called "prometheus" causing a warning to be shown
  • Template the workload type's datasource as '$datasource'
  • Fix CPU and memory panels to filter using the type

Add $cluster filter to Workload network table

  • Add $cluster filter to the "Compute Resources / Workload" network usage table

Fix Compute Resources Node Pods Memory Usage

  • Memory Usage graph filters by $node, which isn't present directly in container_memory_working_set_bytes, but is in the recording rule added by https://github.com/kubernetes-monitoring/kubernetes-mixin/pull/274
  • It appears #274 fixed the Memory Quota table, but missed fixing the usage graph

Merge pull request #285 from dghubble/fix-pods-by-node-memory-usage

Fix Compute Resources Node Pods Memory Usage

Merge pull request #284 from dghubble/master

Fix issues in Compute Resource network panels

add missing tags on network usage dashboards

Merge pull request #287 from alexconlin/add_missing_tags_on_network_dashboards

add missing tags on network usage dashboards - fixes #286

KubePersistentVolumeFullInFourDays - evaluation period from 5 minutes to an hour.

Merge pull request #288 from P6rguVyrst/master

#262 - KubePersistentVolumeFullInFourDays - evaluation period from 5m > 1h

[apps_alerts] - Fix err msg found duplicate series for the match group for KubePodNotReady alert

Signed-off-by: Lord-Y Lord-Y@users.noreply.github.com

change dependency on kausalco/public to grafana/jsonnet-libs as former is a redirect

Merge pull request #294 from paulfantom/grafana-libs-dep

Fix dependency for grafana-builder

[apps_alerts] - Small refactoring of query to manage multiple kube-stats-metrics

Merge pull request #293 from Lord-Y/master

[apps_alerts] - Fix err msg "found duplicate series for the match group"

Created at 3 days ago
pull request opened
Enable dependabot

Update Go modules monthly.

Signed-off-by: SuperQ superq@gmail.com

Created at 3 days ago
create branch
SuperQ create branch superq/dependabot
Created at 3 days ago