DOC-13151 fix zombie 3.x docs issue #814
Merged
Conversation
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
Context: 1) @malarky noted that pages like BAD https://docs.couchbase.com/admin/admin/enterprise-edition.html still exist and are indexed by Google. 2) most recent builds of those docs were replaced with an AWS S3 redirect file that would take you to the DA site, e.g. 404 https://developer.couchbase.com/documentation/server/3.x/admin/pdfs.html 3) ... those links are now unmaintained 404 4) But there are x13 files that were *moved or deleted* and therefore not replaced with those redirection links. Those remain as **zombie* 200 files with bad content under /admin/admin 5) we have in the meantime restored our 3.x archive docs: GOOD https://docs-archive.couchbase.com/docs-3x/admin/Couchbase-intro.html Solution: Step 1: this commit adds nginx rewrites to: * rewrite zombie 200 files without a candidate location to the archive link above * rewrite zombie 200 files that have a good candidate page to a specific page under that tree * rewrite everything else to within that tree under the same path Step 2: once tested, we can expunge these unused files from the /admin/admin/ path in the main bucket.
osfameron requested review from RayOffiah, RichardSmedley, TimLFletcher, ggray-cb and simon-dew 
via @simon-dew, thanks!
osfameron added a commit that referenced this pull request Apr 3, 2025
This reverts commit ad2bedb. 1) This change does not fix the issue it was logged for. In itself, that isn't a huge problem, as it could be iterated on but: 2) I don't have a good way to test the changes beyond deploying to staging. Previously we'd have deployed to `staging` but now (for convenience) those share the same branch, and therefore we can't update one without the other. 3) It's hard to tell if the `sudo update-nginx-rewrites` command succeeded. There's nothing mentioned in the GitHub Workflow log. In the past, when invoked via `ssh` this would have guaranteed that the command ran to completion with a successful exit code. But now, it's run via `aws ssm run`, which only returns that the job is pending. 4) I can't ssh to the production jenkins to examine the state of the nginx there. That's generally not something I want to be doing regularly, but is useful for this kind of case where we're unsure if a change has taken effect. 5) I don't entirely understand the `update-nginx-rewrites` script, so I'm unsure what happens if it fails (does it retain the bad config or revert it? It's been hacked at and commented over some time, including by me, and may not work as specified) 6) `nginx` won't ever *reload* config to a bad config, which is GOOD! BUT... does mean that if a bad config is uploaded, and it doesn't return an error, there is a risk that the bad config was silently uploaded, rejected, but remains on the filesystem... waiting for the time to pouch. 7) If AWS reboot the prod Jenkins for maintenance, the nginx might not come up. This could happen at an unexpected time.
osfameron added a commit that referenced this pull request Apr 8, 2025
This reverts commit b33ce7c. Fixed infrastructure (broken duplicate location removal logic) and tested in infratest.
sarahlwelton pushed a commit that referenced this pull request Jun 23, 2025
* DOC-13151 fix zombie 3.x docs issue Context: 1) @malarky noted that pages like BAD https://docs.couchbase.com/admin/admin/enterprise-edition.html still exist and are indexed by Google. 2) most recent builds of those docs were replaced with an AWS S3 redirect file that would take you to the DA site, e.g. 404 https://developer.couchbase.com/documentation/server/3.x/admin/pdfs.html 3) ... those links are now unmaintained 404 4) But there are x13 files that were *moved or deleted* and therefore not replaced with those redirection links. Those remain as **zombie* 200 files with bad content under /admin/admin 5) we have in the meantime restored our 3.x archive docs: GOOD https://docs-archive.couchbase.com/docs-3x/admin/Couchbase-intro.html Solution: Step 1: this commit adds nginx rewrites to: * rewrite zombie 200 files without a candidate location to the archive link above * rewrite zombie 200 files that have a good candidate page to a specific page under that tree * rewrite everything else to within that tree under the same path Step 2: once tested, we can expunge these unused files from the /admin/admin/ path in the main bucket. * DOC-13151 tweaks to target location via @simon-dew, thanks!
sarahlwelton pushed a commit that referenced this pull request Jun 23, 2025
This reverts commit ad2bedb. 1) This change does not fix the issue it was logged for. In itself, that isn't a huge problem, as it could be iterated on but: 2) I don't have a good way to test the changes beyond deploying to staging. Previously we'd have deployed to `staging` but now (for convenience) those share the same branch, and therefore we can't update one without the other. 3) It's hard to tell if the `sudo update-nginx-rewrites` command succeeded. There's nothing mentioned in the GitHub Workflow log. In the past, when invoked via `ssh` this would have guaranteed that the command ran to completion with a successful exit code. But now, it's run via `aws ssm run`, which only returns that the job is pending. 4) I can't ssh to the production jenkins to examine the state of the nginx there. That's generally not something I want to be doing regularly, but is useful for this kind of case where we're unsure if a change has taken effect. 5) I don't entirely understand the `update-nginx-rewrites` script, so I'm unsure what happens if it fails (does it retain the bad config or revert it? It's been hacked at and commented over some time, including by me, and may not work as specified) 6) `nginx` won't ever *reload* config to a bad config, which is GOOD! BUT... does mean that if a bad config is uploaded, and it doesn't return an error, there is a risk that the bad config was silently uploaded, rejected, but remains on the filesystem... waiting for the time to pouch. 7) If AWS reboot the prod Jenkins for maintenance, the nginx might not come up. This could happen at an unexpected time.
sarahlwelton pushed a commit that referenced this pull request Jun 23, 2025
This reverts commit b33ce7c. Fixed infrastructure (broken duplicate location removal logic) and tested in infratest.
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.
Context:
BAD https://docs.couchbase.com/admin/admin/enterprise-edition.html
still exist and are indexed by Google.
404 https://developer.couchbase.com/documentation/server/3.x/admin/pdfs.html
... those links are now unmaintained 404
But there are x13 files that were moved or deleted and therefore not replaced with those redirection links.
Those remain as *zombie 200 files with bad content under /admin/admin
we have in the meantime restored our 3.x archive docs:
GOOD https://docs-archive.couchbase.com/docs-3x/admin/Couchbase-intro.html
Solution:
Step 1: this commit adds nginx rewrites to:
Step 2: once tested, we can expunge these unused files from the /admin/admin/ path in the main bucket.
NOTE ON Testing
Er, don't currently have a good way to do this, so propose 🤠 (deploy -> test -> rollback).
Happy to put work into improving that if worthwhile, or to schedule this change at the least inconvenient time if not!