fix: dramatically reduce checkpoint memory consumption #2956
Conversation
This file contains 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
…oint
For excessively large tables which do not have sufficient checkpoints or
excessive `action` volume between checkpoints, the
checkpoint code can consume an unreasonable amount of memory.
In the case evaluated there were over a thousand transactions between
transactions, but that resulted in over 2M actions which needed to be
persisted to the checkpoint. This scenario led to 19.9GB of memory
utilized when producing checkpoints for a table which used 3.8GB to
open.
By iteratively processing the buffers which need to be serialized, the
memory is dramatically reduced:
**Before**:
Maximum resident set size (kbytes): 19964728
**After:**
Maximum resident set size (kbytes): 4621648
Sponsored-by: Scribd Inc
Signed-off-by: R. Tyler Croy <rtyler@brokenco.de>
Similar to the prior commit, this improves the iterative nature of
checkpoint creation.
**Before:**
Maximum resident set size (kbytes): 4621648
**After:**
Maximum resident set size (kbytes): 4017132
Sponsored-by: Scribd Inc
Signed-off-by: R. Tyler Croy <rtyler@brokenco.de>
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #2956 +/- ##
==========================================
- Coverage 72.38% 72.35% -0.03%
==========================================
Files 131 131
Lines 40599 40602 +3
Branches 40599 40602 +3
==========================================
- Hits 29387 29378 -9
- Misses 9335 9336 +1
- Partials 1877 1888 +11 ☔ View full report in Codecov by Sentry. |
rtyler
requested review from
wjones127,
roeap,
hntd187 and
ion-elgreco
as code owners
github-merge-queue
bot
removed this pull request from the merge queue due to no response for status checks
ion-elgreco
changed the title
github-merge-queue
bot
removed this pull request from the merge queue due to no response for status checks
alexwilcoxson-rel
pushed a commit
to relativityone/delta-rs
that referenced
this pull request
Nov 8, 2024
Both commits describe the specific fixes, but basically our checkpoint code was collecting too much into memory when it could iterate! :roller_coaster: With a test table: Before `Maximum resident set size (kbytes): 19964728` After `Maximum resident set size (kbytes): 4017132` Sponsored-by: [Scribd Inc](https://tech.scribd.com) --------- Signed-off-by: R. Tyler Croy <rtyler@brokenco.de> (cherry picked from commit c05931a)
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.
Both commits describe the specific fixes, but basically our checkpoint code was collecting too much into memory when it could iterate! 🎢
With a test table:
Before
Maximum resident set size (kbytes): 19964728After
Maximum resident set size (kbytes): 4017132Sponsored-by: Scribd Inc