Repository navigation
v2.6.2
KubeBolt 2.6.2 — right-sizing on a cluster's first day, and Kobi says which role a connector needs
Drop-in over 2.6.1. No schema change and no migration, and the agent stays on
1.4.9. Nothing to do before upgrading: right-sizing keeps its 48-hour
minimum until an admin lowers it.
Right-sizing's minimum history is a setting
Right-sizing recommended nothing until a cluster had two days of usage history.
A cluster connected for a trial or a proof of concept therefore showed "collecting
samples" on its first day, exactly when someone was looking at it.
The minimum is now set in Settings › General › Right-sizing:
- 48 hours by default, the same as before;
- any value from 12 to 168 hours;
- it applies to every cluster and to Kobi;
- it applies on the next read, with no restart.
Recommendations made from less than 48 hours of history are marked as an
early estimate. The panel shows how much history they rest on, for example
"based on 1 day 20h of agent samples · P95 · early estimate". A tooltip explains
that the P95 has not seen two daily peaks yet. Kobi says the same next to its
numbers.
Every "collecting samples" text now names the configured minimum in days and
hours ("needs 2 days", "12h of 1 day 12h collected"), not a fixed "~2 days".
Kobi says which role a cloud connector needs
A person below a connector's minimum role (Who can use it in Kobi, editors and
admins by default) gets no cloud tools. Until now Kobi answered them as if there
were no connector at all.
Kobi now knows which connectors are out of the person's reach and which role
each one needs. When an answer needs those cloud resources, it says so: the
connector, the role, and that an admin can raise the person's role or open the
connector to everyone (Administration › AI (Kobi) › Connectors, Who can use it
in Kobi). Then it goes on with the evidence it can read.
No more burst of 401s every 15 minutes
The access token was renewed only after it had expired. Every 15 minutes, the
next requests in an open tab answered 401 {"error":"invalid or expired token"},
then a refresh, then a retry that succeeded. Nothing failed, but DevTools showed
a column of red rows. A message sent to Kobi in that window showed the 401,
because its chat builds its own request and never retried.
The token is now renewed a minute before it expires, Kobi's chat included. The
401 → refresh → retry path stays for what renewal cannot foresee.
Upgrading
helm upgrade kubebolt oci://ghcr.io/clm-cloud-solutions/kubebolt/helm/kubebolt \
--version 2.6.2 -n kubebolt --reset-then-reuse-valuesUse the namespace you installed into: helm list -A | grep kubebolt shows it.
--reset-then-reuse-values needs Helm 3.14+; passing your values file (-f)
works too. Do not use --reuse-values: it keeps the previous chart's defaults,
pinned image tags included. Step by step, what to expect, rollback, and what
each version asks of you when you skip some:
Upgrade.
The agent stays on 1.4.9: there is no agent release with 2.6.2.
Install
Helm (recommended for Kubernetes)
helm install kubebolt oci://ghcr.io/clm-cloud-solutions/kubebolt/helm/kubebolt \
--namespace kubebolt --create-namespace
kubectl -n kubebolt port-forward svc/kubebolt 3000:80Upgrading an existing install: see Upgrading above and the chart's
Upgrade guide.
Single Binary (download below)
# macOS Apple Silicon
curl -LO https://github.com/clm-cloud-solutions/kubebolt/releases/download/v2.6.2/kubebolt-darwin-arm64
chmod +x kubebolt-darwin-arm64 && mv kubebolt-darwin-arm64 /usr/local/bin/kubebolt
# Linux amd64
curl -LO https://github.com/clm-cloud-solutions/kubebolt/releases/download/v2.6.2/kubebolt-linux-amd64
chmod +x kubebolt-linux-amd64 && sudo mv kubebolt-linux-amd64 /usr/local/bin/kubebolt
kubebolt --kubeconfig ~/.kube/configAll platform binaries are attached below. Verify with sha256sum -c CHECKSUMS.txt.
Docker (single container)
docker run -p 3000:3000 \
-v ~/.kube/config:/kubeconfig:ro -e KUBECONFIG=/kubeconfig \
ghcr.io/clm-cloud-solutions/kubebolt:2.6.2Homebrew (macOS, Linux)
brew install clm-cloud-solutions/tap/kubeboltkubectl krew plugin
kubectl krew index add clm https://github.com/clm-cloud-solutions/krew-index.git
kubectl krew install clm/kubebolt
kubectl kubeboltDocker Compose
git clone https://github.com/clm-cloud-solutions/kubebolt.git
cd kubebolt/deploy && docker compose up -dContainer Images
- Single container (binary + frontend):
ghcr.io/clm-cloud-solutions/kubebolt:2.6.2 - API only:
ghcr.io/clm-cloud-solutions/kubebolt/api:2.6.2 - Web only:
ghcr.io/clm-cloud-solutions/kubebolt/web:2.6.2
Changelog
Features
- feat(kobi): a user below a connector's role floor is told which role it needs ()
Fixes
- fix(rightsizing): the minimum usage history is a setting ()
- fix(web): renew the access token a minute before it expires ()
Other
- Merge pull request #314 from clm-cloud-solutions/develop ()
- Merge pull request #313 from clm-cloud-solutions/chore/release-2.6.2 ()
- chore(release): 2.6.2 ()
- Merge pull request #312 from clm-cloud-solutions/fix/rightsizing-min-history ()
- Merge pull request #311 from clm-cloud-solutions/fix/web-token-renew-ahead ()
- Merge pull request #310 from clm-cloud-solutions/fix/kobi-connector-floor-notice ()