[SERVER-63210] Verify spilling logic in HashAgg works for $lookup HJ plans Created: 02/Feb/22 Updated: 16/Feb/22 Resolved: 16/Feb/22 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Task | Priority: | Major - P3 |
| Reporter: | Eric Cox (Inactive) | Assignee: | Anna Wawrzyniak |
| Resolution: | Won't Fix | Votes: | 0 |
| Labels: | sbe | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Participants: |
| Description |
|
In PM-2267 spilling to disk in HashAgg stage was implemented to support $group queries that hit the memory limit and would spill to a WT temp store when allowDiskUse=true. Given that we will use the HashAgg stage to build a hash table for the foreign side and accumulate the set of documents that satisfy the eq predicate between the local/foreign side keys the current spilling implementation might just work for our use-case. The owner of this ticket should exercise the spilling implementation in HashAgg and see if it is suitable for HJ plans. If changes need to be made to support HJ plans we should do so under this work. |
| Comments |
| Comment by Eric Cox (Inactive) [ 16/Feb/22 ] |
|
We decided to implement HJ in a new sbe PlanStage so it will require a new spilling implementation. |