DAOS-19292 doc: Remove unsafe PMEM_NO_FLUSH guidance#18703
Open
knard38 wants to merge 1 commit into
Open
Conversation
docs/admin/performance_tuning.md instructed administrators to manually export PMEM_NO_FLUSH=1 on 3rd Gen Intel Xeon Scalable (ICX) systems to take advantage of eADR. This bypasses PMDK's own safe, runtime detection (pmem_has_auto_flush()) of whether the platform's persistence domain actually covers CPU caches. CPU generation is a necessary but not sufficient condition for eADR: BIOS configuration, OEM platform variants, and hardware health also affect whether eADR is actually active. Following this note on a system that matches the CPU generation but does not have eADR working would silently risk data loss on the next power failure. Remove the note entirely. No DAOS source code sets PMEM_NO_FLUSH; this documentation note was the only occurrence in the repository. Doc-only: true Signed-off-by: Cedric Koch-Hofer <cedric.koch-hofer@hpe.com>
|
Ticket title is 'Remove unsafe |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
docs/admin/performance_tuning.mdinstructed administrators to manually exportPMEM_NO_FLUSH=1on 3rd Gen Intel® Xeon® Scalable (ICX) systems to "take advantage of" the extended asynchronous DRAM refresh (eADR) feature. This note is removed because it bypasses PMDK's own safe, built-in runtime detection of whether the platform's persistence domain actually covers CPU caches.PMEM_NO_FLUSH=1unconditionally forceslibpmemto never issueCLFLUSH/CLFLUSHOPT/CLWBinstructions, regardless of whether the platform truly guarantees the CPU-cache persistence domain. CPU generation alone is a necessary but not sufficient condition for eADR: BIOS configuration, OEM platform implementation, and hardware health also determine whether eADR is actually active. An administrator who follows this note on a system that matches the CPU generation but does not have eADR enabled/working (e.g. disabled in BIOS, unsupported OEM variant, or a genuine ADR/eADR hardware fault) would silently risk data loss on the next power failure, since PMDK would skip mandatory cache-flush instructions while believing them unnecessary.PMDK already exposes a safe, automatic alternative,
pmem_has_auto_flush(), which queries the platform's actual persistence-domain capability at runtime. PMDK's ownlibpmem.7manpage explicitly frames manual overrides likePMEM_NO_FLUSHas testing-only ("largely intended for testing and are not normally required").A repository-wide search confirms no DAOS source code sets
PMEM_NO_FLUSHanywhere; this documentation note was the only occurrence in thedaos-stack/daosrepository, so no other doc, script, or test config needs to be updated.Steps for the author:
After all prior steps are complete: