HPCC-30564 Prevent pending write externals being flushed from cache. #17953
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
jakesmith
force-pushed
the
HPCC-30564-extcache-fix
branch
2 times, most recently
from
75938dd to
3a3d94d
Compare
jakesmith
commented
Oct 25, 2023
| @@ -1956,6 +1967,8 @@ class CExtCache | |||
| if (mapIt != extTable.end()) | |||
| { | |||
| auto listIter = mapIt->second; | |||
| Linked<CTransactionItem> &existingItem = *listIter; | |||
| cachedSz -= existingItem->ext.dataLength; | |||
There was a problem hiding this comment.
@ghalliday - I spotted this in some late testing. Basically it would mean that although the cache was low, the cachedSz was left high, resulting in repeated spurious premature attempts to purge the ext. cache. It would mean that the external cache was effectively a 1-element cache after running for some time.
ghalliday
reviewed
Oct 25, 2023
dali/base/dasds.cpp
Outdated
| // pending on write. | ||
| unsigned extCacheSizeMB = config.getPropInt("@extCacheSizeMB", defaultExtCacheSizeMB); | ||
| unsigned deltaTransactionMaxMemMB = config.getPropInt("@deltaTransactionMaxMemMB", defaultDeltaMemMaxMB); | ||
| if (extCacheSizeMB && deltaTransactionMaxMemMB && (extCacheSizeMB < deltaTransactionMaxMemMB)) |
There was a problem hiding this comment.
Should this be <= rather than <?
There was a problem hiding this comment.
yes, it should, I'll change and squash that minor change in.
Flushing some externals from the external cache (when full), whilst the transactions are still pending to be written could cause subsequent ext. cache lookups to attempt to look to disk for the external, and fail with "Missing external file ". Also fix a bug with cachedSz. Transactions that removed externals, were being removed from the cache, but their size was not deducted from cachedSz. Consequently, over time, the extCache would falsely appear full, meaning it was purged prematurely. Signed-off-by: Jake Smith <[email protected]>
jakesmith
force-pushed
the
HPCC-30564-extcache-fix
branch
from
3a3d94d to
588dccf
Compare
ghalliday
approved these changes
Oct 26, 2023
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.
Flushing some externals from the external cache (when full), whilst the transactions are still pending to written could cause subsequent ext. cache lookups to attempt to look to disk for the external, and fail with "Missing external file ".
Type of change:
Checklist:
Smoketest:
Testing: