opt(restore): Sort the buffer before spinning the writeToDisk goroutine #7984

ahsanbarkati · 2021-08-12T18:16:57Z

Sort the buffer before-hand instead of sorting it in the goroutine used for
writing the buffer to disk. The writeToDisk goroutines are throttled and making it
expensive causes other goroutines to block.

This change significantly improves restore map phase.

This change is

manishrjain

Awesome change!

Reviewed 1 of 1 files at r1, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @ahsanbarkati)

…ne (#7984) Sort the buffer beforehand instead of sorting it in the goroutine used for writing the buffer to disk. The writeToDisk goroutines are throttled and making it expensive causes other goroutines to block. This change significantly improves the restore map phase. (cherry picked from commit 1966245)

…ne (#7984) (#7996) Sort the buffer beforehand instead of sorting it in the goroutine used for writing the buffer to disk. The writeToDisk goroutines are throttled and making it expensive causes other goroutines to block. This change significantly improves the restore map phase. (cherry picked from commit 1966245) Co-authored-by: Ahsan Barkati <ahsan@dgraph.io>

This commit is a major rewrite of backup and online restore code. It used to use KVLoader in badger. Now it instead uses StreamWriter that is much faster for writes. cherry-pick PR #7753 following commits are cherry-picked (in reverse order): * opt(restore): Sort the buffer before spinning the writeToDisk goroutine (#7984) (#7996) * fix(backup): Fix full backup request (#7932) (#7933) * fix: fixing graphql schema update when the data is restored + * fix(restore): return nil if there is error (#7899) skipping /probe/graphql from audit (#7925) * Don't ban namespace in export_backup * reset the kv.StreamId before sending to stream writer (#7833) (#7837) * fix(restore): Bump uid and namespace after restore (#7790) (#7800) * fix(ee): GetKeys should return an error (#7713) (#7797) * fix(backup): Free the UidPack after use (#7786) * fix(export-backup): Fix double free in export backup (#7780) (#7783) * fix(lsbackup): Fix profiler in lsBackup (#7729) * Bring back "perf(Backup): Improve backup performance (#7601)" * Opt(Backup): Make backups faster (#7680) * Fix s3 backup copy (#7669) * [BREAKING] Opt(Restore): Optimize Restore's new map-reduce based design (#7666) * Perf(restore): Implement map-reduce based restore (#7664) * feat(backup): Merge backup refactoring * Revert "perf(Backup): Improve backup performance (#7601)"

This commit is a major rewrite of backup and online restore code. It used to use KVLoader in badger. Now it instead uses StreamWriter that is much faster for writes. cherry-pick PR #7753 following commits are cherry-picked (in reverse order): * opt(restore): Sort the buffer before spinning the writeToDisk goroutine (#7984) (#7996) * fix(backup): Fix full backup request (#7932) (#7933) * fix: fixing graphql schema update when the data is restored + skipping /probe/graphql from audit (#7925) * fix(restore): return nil if there is error (#7899) * Don't ban namespace in export_backup * reset the kv.StreamId before sending to stream writer (#7833) (#7837) * fix(restore): Bump uid and namespace after restore (#7790) (#7800) * fix(ee): GetKeys should return an error (#7713) (#7797) * fix(backup): Free the UidPack after use (#7786) * fix(export-backup): Fix double free in export backup (#7780) (#7783) * fix(lsbackup): Fix profiler in lsBackup (#7729) * Bring back "perf(Backup): Improve backup performance (#7601)" * Opt(Backup): Make backups faster (#7680) * Fix s3 backup copy (#7669) * [BREAKING] Opt(Restore): Optimize Restore's new map-reduce based design (#7666) * Perf(restore): Implement map-reduce based restore (#7664) * feat(backup): Merge backup refactoring * Revert "perf(Backup): Improve backup performance (#7601)"

This commit is a major rewrite of backup and online restore code. It used to use KVLoader in badger. Now it instead uses StreamWriter that is much faster for writes. cherry-pick PR #7753 following commits are cherry-picked (in reverse order): * opt(restore): Sort the buffer before spinning the writeToDisk goroutine (#7984) (#7996) * fix(backup): Fix full backup request (#7932) (#7933) * fix: fixing graphql schema update when the data is restored + * fix(restore): return nil if there is error (#7899) skipping /probe/graphql from audit (#7925) * Don't ban namespace in export_backup * reset the kv.StreamId before sending to stream writer (#7833) (#7837) * fix(restore): Bump uid and namespace after restore (#7790) (#7800) * fix(ee): GetKeys should return an error (#7713) (#7797) * fix(backup): Free the UidPack after use (#7786) * fix(export-backup): Fix double free in export backup (#7780) (#7783) * fix(lsbackup): Fix profiler in lsBackup (#7729) * Bring back "perf(Backup): Improve backup performance (#7601)" * Opt(Backup): Make backups faster (#7680) * Fix s3 backup copy (#7669) * [BREAKING] Opt(Restore): Optimize Restore's new map-reduce based design (#7666) * Perf(restore): Implement map-reduce based restore (#7664) * feat(backup): Merge backup refactoring * Revert "perf(Backup): Improve backup performance (#7601)"

This commit is a major rewrite of backup and online restore code. It used to use KVLoader in badger. Now it instead uses StreamWriter that is much faster for writes. cherry-pick PR #7753 following commits are cherry-picked (in reverse order): * opt(restore): Sort the buffer before spinning the writeToDisk goroutine (#7984) (#7996) * fix(backup): Fix full backup request (#7932) (#7933) * fix: fixing graphql schema update when the data is restored + skipping /probe/graphql from audit (#7925) * fix(restore): return nil if there is error (#7899) * Don't ban namespace in export_backup * reset the kv.StreamId before sending to stream writer (#7833) (#7837) * fix(restore): Bump uid and namespace after restore (#7790) (#7800) * fix(ee): GetKeys should return an error (#7713) (#7797) * fix(backup): Free the UidPack after use (#7786) * fix(export-backup): Fix double free in export backup (#7780) (#7783) * fix(lsbackup): Fix profiler in lsBackup (#7729) * Bring back "perf(Backup): Improve backup performance (#7601)" * Opt(Backup): Make backups faster (#7680) * Fix s3 backup copy (#7669) * [BREAKING] Opt(Restore): Optimize Restore's new map-reduce based design (#7666) * Perf(restore): Implement map-reduce based restore (#7664) * feat(backup): Merge backup refactoring * Revert "perf(Backup): Improve backup performance (#7601)"

Fix restore bottleneck

7ea5db4

ahsanbarkati requested a review from manishrjain as a code owner August 12, 2021 18:16

manishrjain approved these changes Aug 12, 2021

View reviewed changes

ahsanbarkati merged commit 1966245 into master Aug 12, 2021

ahsanbarkati deleted the ahsan/fix-restore-bottleneck branch August 12, 2021 20:27

danielmai mentioned this pull request Aug 24, 2021

opt(restore): Sort the buffer before spinning the writeToDisk goroutine (#7984) #7996

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

opt(restore): Sort the buffer before spinning the writeToDisk goroutine #7984

opt(restore): Sort the buffer before spinning the writeToDisk goroutine #7984

ahsanbarkati commented Aug 12, 2021 •

edited by manishrjain

Loading

manishrjain left a comment

opt(restore): Sort the buffer before spinning the writeToDisk goroutine #7984

opt(restore): Sort the buffer before spinning the writeToDisk goroutine #7984

Conversation

ahsanbarkati commented Aug 12, 2021 • edited by manishrjain Loading

manishrjain left a comment

Choose a reason for hiding this comment

ahsanbarkati commented Aug 12, 2021 •

edited by manishrjain

Loading