Sparse without dups unordered uses indexreader #3347
Conversation
This fixes a mismatch between the serialization path and the query strategy decision code that resulted in the serialization path incorrectly trying to load the base reader state instead of the indexed reader state.
|
This pull request has been linked to Shortcut Story #19264: Sparse Unordered Without Dups Reader is indexed for serialization. |
|
Is this related to the change that was made in 2.10 to revert back to legacy for the sparse global order reader in case we have a query condition? If so, this change was merge directly into 2.10. This will need to be merged directly there as well since nothing was changed in dev yet for this. |
The goal is that we correctly serialize the right reader. Since we do not have a centralized place for making the determination about reader things got out of sync. This is a regression and we need to get this into 2.10.2 as soon as possible as this is blocking a lot of work. If you want to handle things differently I'm open but we have to have tbis merged this morning and shipped. I wanted this shipped yesterday but didn't get a review. |
|
Merge as is, I'll fix properly for 2.11... |
Shelnutt2
deleted the
sethshelnutt/sc-19264/sparse-unordered-without-dups-reader-is-indexed
branch
|
The backport to To backport manually, run these commands in your terminal: # Fetch latest updates from GitHub
git fetch
# Create a new working tree
git worktree add .worktrees/backport-release-2.10 release-2.10
# Navigate to the new working tree
cd .worktrees/backport-release-2.10
# Create a new branch
git switch --create backport-3347-to-release-2.10
# Cherry-pick the merged commit of this pull request and resolve the conflicts
git cherry-pick --mainline 1 5e02006cc8beec669a1c6ccf9f1a06ea3fd231b2
# Push it to GitHub
git push --set-upstream origin backport-3347-to-release-2.10
# Go back to the original working tree
cd ../..
# Delete the working tree
git worktree remove .worktrees/backport-release-2.10Then, create a pull request where the |
This fixes a mismatch between the serialization path and the query strategy decision code that resulted in the serialization path incorrectly trying to load the base reader state instead of the indexed reader state.
This fixes a mismatch between the serialization path and the query strategy decision code that resulted in the serialization path
incorrectly trying to load the base reader state instead of the indexed reader state.
This was causing a
Error: Invalid Layouterror in de-serializing.A followup story has been opened about centralizing the logic for handling query strategies.
TYPE: BUG
DESC: Fix issue with sparse unordered without duplicates query deserialization to use Indexed Reader