Fix/TS expiration handling without network #2937
Open
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
carpawell
changed the title
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #2937 +/- ##
==========================================
+ Coverage 23.90% 24.01% +0.10%
==========================================
Files 776 773 -3
Lines 45708 45597 -111
==========================================
+ Hits 10928 10949 +21
+ Misses 33920 33787 -133
- Partials 860 861 +1 ☔ View full report in Codecov by Sentry. |
carpawell
requested review from
roman-khimov and
cthulhu-rider
as code owners
carpawell
force-pushed
the
fix/TS-expiration-handling-without-network
branch
from
8831c47
to
8314653
Compare
It takes too much of network/disk to handle tombstones expiration and removal, so it is better to store it when tombstone is being indexed in metabase. Additional little-endian expiration suffix was added to graveyard values. Metabase version was increased as it is a non-compatible change. Relates #2929. Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
They do not differ from the other objects (e.g. locks do). Initial logic has changed much, graveyard now allows to handle expired tombstones marks (do not confuse it with the lists of regular indexes) independently, while disk can be cleared with the other types of object. Also, add tests for it. Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Graveyard now has tombstone expiration marks in epochs, there is no need to use any network requests, just drop records if an epoch is big enough. Closes #2929. Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
It may or may not index tombstone's expiration in graveyard. Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
carpawell
force-pushed
the
fix/TS-expiration-handling-without-network
branch
from
8314653
to
152150e
Compare
Migration can be done automatically, let admin's life be a better thing. TS expiration is taken as the current epoch + 5. It is not critical if TS will live more. In practice, side effects _may_ be seen only the first 5 hours after an update, and only returned status code _may_ differ. Signed-off-by: Pavel Karpy <carpawell@nspcc.ru>
| @@ -49,7 +58,7 @@ func checkVersion(tx *bbolt.Tx, initialized bool) error { | |||
| return nil | |||
| } | |||
|
|
|||
| func updateVersion(tx *bbolt.Tx, version uint64) error { | |||
| func (db *DB) updateVersion(tx *bbolt.Tx, version uint64) error { | |||
There was a problem hiding this comment.
Why making it a method if db is never used?
| defer db.modeMtx.RUnlock() | ||
|
|
||
| if db.mode.NoMetabase() { | ||
| return -1, ErrDegradedMode |
There was a problem hiding this comment.
dont see any sense in negative return, can be zero?
| if binary.LittleEndian.Uint64(v[addressKeySize:]) < epoch { | ||
| err := c.Delete() | ||
| if err != nil { | ||
| return fmt.Errorf("cleared %d TS marks in %d epoch and got error: %w", counter, epoch, err) |
There was a problem hiding this comment.
id add context to the error outside db handler
| k, v := c.First() | ||
|
|
||
| for k != nil { | ||
| if binary.LittleEndian.Uint64(v[addressKeySize:]) < epoch { |
There was a problem hiding this comment.
i'd still check v length explicitly to avoid unclear panic, especially since the suffix was not always there
| var counter int | ||
|
|
||
| err := db.boltDB.Update(func(tx *bbolt.Tx) error { | ||
| bkt := tx.Bucket(graveyardBucketName) |
| return res, fmt.Errorf("decode tombstone address from value: %w", err) | ||
| } | ||
|
|
||
| if len(v) == addressKeySize+8 { |
There was a problem hiding this comment.
> case should be either erroneous or no-problem, silence is bad
|
|
||
| func migrateFrom2Version(db *DB, tx *bbolt.Tx) error { | ||
| tsExpiration := db.epochState.CurrentEpoch() + objectconfig.DefaultTombstoneLifetime | ||
| bkt := tx.Bucket(graveyardBucketName) |
|
|
||
| for k, v := c.First(); k != nil; k, v = c.Next() { | ||
| newVal := make([]byte, addressKeySize, addressKeySize+8) | ||
| copy(newVal, v) |
There was a problem hiding this comment.
imo we should assert len(v) == addressKeySize
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.
No description provided.