Repository navigation
docs: record proTES configuration - #179
Merged
Merged
Conversation
Record the effective configuration of the ELIXIR On-Cloud's proTES on its operator page: deployment, token issuer and audience, group mapping and the added admin rules, TES services and plugins. Mark the page verified against the live service info and configuration, and name ECP_CLN:ADMIN next to EOC-Compute in the access row. Refs elixir-cloud-aai/generic-issues#31 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
Reviewer's GuideDocumentationRecord the verified proTES deployment and access configuration, including its group mapping, TES backend, plugins, and follow-up for service-token support.
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
Documentation build overview
|
uniqueg
marked this pull request as ready for review
November 18, 2026 05:12
Contributor
Linked issuesChecked at commit 1882e73 #31 chore: configure proTES for ELIXIR On-Cloud · 0 of 6 met
|
Contributor
There was a problem hiding this comment.
Hey - I've found 4 issues
Prompt for AI Agents
Please address the comments from this code review:
## Individual Comments
### Comment 1
<location path="docs/operators/services/protes.md" line_range="69" />
<code_context>
+| Role `admin` | `urn:geant:lifescience-ri.eu:group:elixir:ECP_CLN:ADMIN#aai.lifescience-ri.eu`: the permissions API (`/admin/access-control/permissions`, on the public address) and the provenance of all tasks; three added rules (`admins-create-list-tasks`, `admins-list-plugins`, `admins-get-plugins`) give it the rights of `user` too |
+| Other tokens | valid tokens without either group get `403 Forbidden` on tasks and plugins; the service info is public |
+| Own tasks | only the user who created a task can read and cancel it; admins can read the provenance of any task, not the task |
+| TES services | the TESK for tests at CSC, without authentication (`auth: none`) |
+| Plugins | none configured (the service info lists no capabilities) |
+
</code_context>
<issue_to_address>
**Tasks cannot reach live Node backends**
When proTES is expected to serve production tasks through live Node backends, the documented TES configuration lists only the unauthenticated CSC test TESK and says no Node TES service has been added, leaving live production backends without configuration or documented authentication modes; operators see only the sandbox service.
Add each live Node TES backend to the proTES configuration and document its authentication mode.
</issue_to_address>
### Comment 2
<location path="docs/operators/services/protes.md" line_range="77-80" />
<code_context>
+`ECP_CLN:ADMIN` assigned to its facility and its client ID added to the
+audience (see [LS Login](../ls-login.md)).
+
+proTES doesn't use a client of its own yet. Once it calls the TES services
+with its own LS Login token, the TESK at CSC moves to an address that
+requires tokens and its entry here changes from `auth: none` to
+`client_credentials` ([generic-issues#111][gi-111]).
+
+[gi-111]: <https://github.com/elixir-cloud-aai/generic-issues/issues/111>
</code_context>
<issue_to_address>
**Own client still absent**
The issue asks for a proTES client-credentials client and deployment Secret. The page instead says proTES has no client of its own yet and defers that work to generic-issues#111.
</issue_to_address>
### Comment 3
<location path="docs/operators/services/protes.md" line_range="40" />
<code_context>
| Address | TES API at `https://protes.rahtiapp.fi/ga4gh/tes/v1` (the host itself has no page) |
| Version | proTES 0.6.0 |
-| Access | LS Login access tokens of members of the group `ECP_CLN:EOC-Compute`, issued to a client that it accepts, such as Krini's (see [Researchers](../../researchers/index.md#before-you-start)) |
+| Access | LS Login access tokens of members of the group `ECP_CLN:EOC-Compute` or `ECP_CLN:ADMIN`, issued to a client that it accepts, such as Krini's (see [Researchers](../../researchers/index.md#before-you-start)) |
| Environment | `test`: its service info calls it "proTES sandbox (BioHackathon 2026)" |
| TES services | one, listed in its configuration: the TESK for tests at CSC (`csc-tesk-noauth.rahtiapp.fi`), which accepts tasks without a token |
</code_context>
<issue_to_address>
**Admins still told to join compute**
When a member of `ECP_CLN:ADMIN` without `ECP_CLN:EOC-Compute` follows the researcher start instructions to use proTES, the researchers start page still presents `ECP_CLN:EOC-Compute` as a prerequisite, so an admin-only member following it is told to request a group that this proTES access row says they do not need.
Clarify that `ECP_CLN:ADMIN` is an alternative for proTES access, or scope the compute-group prerequisite to services that require it.
</issue_to_address>
### Comment 4
<location path="docs/operators/services/protes.md" line_range="50" />
<code_context>
-services that it uses. The rest of its configuration, such as its plugins
-and how it authenticates at its TES services, is tracked in
-[generic-issues#31][gi-31].
+services that it uses.
-[gi-31]: <https://github.com/elixir-cloud-aai/generic-issues/issues/31>
</code_context>
<issue_to_address>
**Operators follow stale configuration link**
When a node operator follows the onboarding tutorial to find proTES configuration, the onboarding tutorial still directs readers to generic-issues#31 even though this page now records the configuration and the PR closes that issue, so operators land on a stale source instead of the maintained operator page.
Update the tutorial's proTES configuration pointer to this operator page and remove the outdated claim that the configuration is tracked in generic-issues#31.
Also at `docs/operators/services/protes.md:52`.
</issue_to_address>Sourcery assessment
Approval pending. 4 findings to address first.
Blocking findings: docs/operators/services/protes.md:69, docs/operators/services/protes.md:80, docs/operators/services/protes.md:40, docs/operators/services/protes.md:50
The group mapping matches the EOC-Compute and ADMIN subgroup values, not the parent ECP_CLN value, which LS Login releases only if the parent group is assigned to the client's facility (generic-issues#46). Addresses a Sourcery review thread. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
14 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Records the effective configuration of the ELIXIR On-Cloud's proTES on its operator page (
docs/operators/services/protes.md), as decided for generic-issues#31: deployment (chart and image of the v0.6.0 release on Rahti), LS Login as the only issuer, the accepted clients (Krini, Cloud Registry UI), groups from userinfo, the group mapping (ECP_CLN:EOC-Compute->user,ECP_CLN:ADMIN->adminplus the three added rules that let admins run tasks), the TES service (CSC test TESK,auth: none) and plugins (none). The page no longer points to generic-issues#31 for this; it points to generic-issues#111 for the switch to proTES' own service token.Also: the access row now names
ECP_CLN:ADMINnext toECP_CLN:EOC-Compute, andlast_verifiedis set to 2026-11-18, as every fact of the page was checked against its source (live service info, the deployed configuration read withoc get(no Secrets), the proTES v0.6.0 docs).Light lane: docs-only change to one page, no code, CI or configuration.
Verification
GET https://protes.rahtiapp.fi/ga4gh/tes/v1/service-info: version 0.6.0, TES 1.1.0, environmenttest, one backendhttps://csc-tesk-noauth.rahtiapp.fi, no capabilities.Deployed config overlay (ConfigMap
protes-config) and images (oc get deploy,statefulset) match the table.python scripts/check_names.py: ok.pytest: 356 passed, 100% coverage.DOCS_REGISTRY_OFFLINE=1 mkdocs build --strict: ok. The online build couldn't run locally:https://elixir-on.cloud/ga4gh/registry/v1/servicestimed out from this machine; CI runs the online build.After merging
mainand the Sourcery fix (1882e73): CI green (CI passed, Docs, Test, PR title, CodeQL), Read the Docs build succeeded, Sourcery check passed.Declined review findings
Light lane, self-reviewed. Sourcery threads declined, with replies in each thread:
ECP_CLN:ADMINisn't requested there.EOC-ComputeandADMINmembers run tasks. The page now says so (1882e73).Open questions
None.
Closes elixir-cloud-aai/generic-issues#31
🤖 Generated with Claude Code
Summary by Sourcery
Document the verified production configuration and access model for the ELIXIR On-Cloud proTES service.
Enhancements:
Documentation: