Releases: Altinity/clickhouse-backup
Releases · Altinity/clickhouse-backup
2.4.9
2.4.8
2.4.7
2.4.6
v2.4.6
IMPROVEMENTS
- make 'kopia' custom scripts really increment fix 781
- add
force_http
and improve retries in GCS upload 784, thanks @minguyen9988
BUG FIXES
2.4.5
v2.4.5
BUG FIXES
- add aws.LogResponse for
S3_DEBUG
(affected 2.4.4+ versions)
2.4.4
v2.4.4
BUG FIXES
- remove aws.LogResponseWithBody for
S3_DEBUG
to avoid too many logs (affected 2.4.0+ versions)
2.4.3
2.4.2
v2.4.2
IMPROVEMENTS
- add
cpu_nice_priority
andio_nice_priority
togeneral
config section, which allow us to throttle CPU and IO usage for the wholeclickhouse-backup
process, fix 757
BUG FIXES
- fix restore for object disk frozen_metadata.txt fix 752
- fix more corner cases for
check_parts_columns: true
, fix 747 - fix applying macros to s3 endpoint in object disk during restore embedded backups, fix 750
- rewrite GCS clients pool, set default GCS_CLIENT_POOL_SIZE as max(upload_concurrency, download_concurrency) * 3 to avoid stuck, fix 753, thanks @minguyen-jumptrading
2.4.1
v2.4.1
IMPROVEMENTS
- switch to go-1.21
- add clickhouse-server:23.8 for integration and testflows tests
- add
FTP_SKIP_TLS_VERIFY
config option fix 742
BUG FIXES
- fix calculation part size for
S3
and buffer size forAzure
to avoid errors for upload big files, fix 739 thanks @rodrigargar - fix GetFileReader for SSE encryption in
S3
, again fix 709
2.4.0
v2.4.0
IMPROVEMENTS
- first implementation for properly backup S3/GCS/Azure disks, support server-side copy to back up bucket during
clickhouse-backup
create and duringclickhouse-backup restore
, requires addobject_disk_path
tos3
,gcs
,azblob
section, fix 447 - Implementation blacklist for table engines during backup / download / upload / restore 537
- restore RBAC / configs, refactoring restart clickhouse-server via
sql:SYSTEM SHUTDOWN
orexec:systemctl restart clickhouse-server
, add--rbac-only
and--configs-only
options tocreate
,upload
,download
,restore
command. fix 706 - Backup/Restore RBAC related objects from Zookeeper via direct connection to zookeeper/keeper, fix 604
- Add
SHARDED_OPERATION_MODE
option, to easy create backup for sharded cluster, available valuesnone
(no sharding),table
(table granularity),database
(database granularity),first-replica
(on the lexicographically sorted first active replica), thanks @mskwon, fix 639, fix 648 - Add support for
compression_format: none
for upload and download backups created with--rbac
/--rbac-only
or--configs
/--configs-only
options, fix 713 - Add support for s3
GLACIER
storage class, when GET return error, then, it requires 5 minutes per key and restore could be slow. UseGLACIER_IR
, it looks more robust, fix 614 - restore functions via
CREATE OR REPLACE
for more atomic behavior - prepare to Make ./tests/integration/ test parallel fix 721
- touch
/var/lib/clickhouse/flags/force_drop_table
before every DROP TABLE execution, fix 683 - add support connection pool for google cloud storage,
GCS_CLIENT_POOL_SIZE
, fix 724
BUG FIXES
- fix possible create backup failures during UNFREEZE not exists tables, affected 2.2.7+ version, fix 704
- fix too strict
system.parts_columns
check when backup create, exclude Enum and Tuple (JSON) and Nullable(Type) vs Type corner cases, fix 685, fix 699 - fix
--rbac
behavior when /var/lib/clickhouse/access not exists - fix
skip_database_engines
behavior - fix
skip_databases
behavior during restore for corner casedb.prefix*
and corner cases when conflict with--tables="*pattern.*"
, fix 663 - fix S3 head object Server Side Encryption parameters, fix 709