v2.11.1.0 - December 15, 2021
Build: 2.11.1.0-b305
Downloads
macOS Linux x86 Linux ARMDocker
docker pull yugabytedb/yugabyte:2.11.1.0-b305
New Features
Yugabyte Platform
- [PLAT-526] [UI] UI support for xCluster replication management phase-1
- [PLAT-529] GFlags validation APIs
Database
- [6639] [YSQL] Add ALTER TABLE SET TABLESPACE
- [8242] [DST] Enable automatic tablet splitting by default
- [9468] [YSQL] [Part-1] READ COMMITTED isolation level
- [9984] Support arm64 build on macOS
- [10563] [xCluster] Create Rename Replication Command
- [10655] [YSQL] Add
--postmaster_cgroup
gflag to tserver - [10671] [yugabyted] use env var YB_DISABLE_CALLHOME
Improvements
Yugabyte Platform
- [8517] [6302] [5355] Auto populate provider dropdown, remove parenthesis around range partition and adjust DB version margin (#10229)
- [8517] Add handling for provider input in edit universe scenario
- [PLAT-506] Allow custom port in AWS host base
- [PLAT-1773] Auto refresh of metrics page with intervals dropdown
- [PLAT-1783] Add loading icon for universe list page
- [PLAT-1923] [PLAT-1924] Ability to create universes with AZs count > RF + definition of default region
- [PLAT-1931] Add Prometheus links to metrics panels
- [PLAT-1941] Add status quick filter and table actions
- [PLAT-1944] Allow customers to modify platform log rollover and retention settings
- [PLAT-1950] Collect additional metrics
- [PLAT-1957] Check ports availability in preflight check
- [PLAT-1993] [UI] Show backup config name in list of backups
- [PLAT-2004] Add leader blacklisting to UpgradeTaskBase
- [PLAT-2061] Add PagerDuty and WebHook channels
- [PLAT-2063] YSQL connections count, Leaderless tablets, Under-replicated tablets, Missing master leader alerts
- [PLAT-2064] DB Overload alerts
- [PLAT-2108] NFS backup with Encryption at rest enabled
- [PLAT-2143] [UI] Add an optional field AWS KMS Endpoint while creating KMS config
- [PLAT-2154] Add UI support for new log search API endpoint
- [PLAT-2173] Task Improvement - implement ways to save and read states changed by subtasks
- [PLAT-2195] Add support for multi-universe filter alert count endpoint
- [PLAT-2227] Added a runtime config setting for disabling all the cert validations
- [PLAT-2229] Retrieve YB version via ssh/kubectl during health check
- [PLAT-2237] Add leader blacklisting to stop node in universe
- [PLAT-2244] Update instance creation and listing to accept and return node UUID and Universe UUID.
- [PLAT-2261] Enhancements to /logs UI endpoint
- [PLAT-2303] Change Prometheus link icon in Metrics Page
- [PLAT-2306] Add UI support for GCP user tags
- [PLAT-2345] Show metrics for paused universes
- [PLAT-2348] Increase the configuration name max size to 100
- [PLAT-2361] Filter and sort alert configurations on various fields + return alert count in page API
- [PLAT-2467] [UI] XCluster UI Changes
- Update certs to use certLabel as CN
Database
- [1127] [YSQL] Improve collation upgrade performance.
- [1127] [YSQL] Use attribute collation for SPLIT AT clause
- [1979] Support PG11 text search configuration
- [5395] [docdb] Add metrics for master async tasks
- [7809] [YSQL] Add session flag to enable non-transactional writes on COPY
- [8597] [YSQL] Changing Catalog Version Mismatch in Postgres layer as retryable error
- [10158] [Geo] Selecting different txn status table based on local vs global requirements
- [10190] [YSQL] create and use the pg_yb_tablegroup table
- [10159] [Geo] Added global_transaction session variable
- [10352] [YSQL] Import Invalidate partitions of table being attached/detached
- [10358] [xCluster] [Tablet-splitting] Handle consumer side splits
- [10428] [DocDB] Improve logging for SST file expiration
- [10388] [YSQL] Import Fix pg_dump for disabled triggers on partitioned tables
- [10431] [YSQL] Import pg_dump: fix mis-dumping of non-global default privileges.
- [10433] [YSQL] Catalog cache refresh improvements
- [10435] [DocDB] Display split depth in the master Tables UI
- [10503] [DocDB] Add GFlag to trust value-level TTL metadata during file expiration.
- [10543] [YCQL] Allow NULL in filtering clauses WHERE/IF inside IN/NOT IN operators in SELECT statement.
- [10576] [YSQL] Import Reject extraneous data after SSL or GSS encryption handshake.
- [10579] [YSQL] Import libpq: reject extraneous data after SSL or GSS encryption handshake.
- [10586] [YSQL] Reduce number of reads from the pg_yb_catalog_version table on staring new connection
- [10604] [YCQL] Handle DeleteIndex while backfilling
- Update cassandra driver version to 3.10.3-yb-2 towards vulnerabilities fixes.
Bug Fixes
Yugabyte Platform
- [PLAT-1643] On-prem provider configuration, "Desired Home Directory" can't be equal to the first mount paths item
- [PLAT-1678] Multitable backup fails with NPE
- [PLAT-1702] [UI] Azure YW provider does not allow port customization
- [PLAT-1795]
/home/yugabyte
not getting cleared out with a custom home dir for onprem universe - [PLAT-1891] [Backup] [IAM-Platform] Backup is hanging for universe with read replicas on IAM-enabled platform
- [PLAT-1948] [UI] Delete release should call delete API, not update release API
- [PLAT-2088] Some primary cluster fields not inherited properly into add read replica page
- [PLAT-2098] fix: Certificate details page shows 'invalid date' for certificate start and expiration fields on Safari Browser only
- [PLAT-2099] Remove taskUUID from audit entry while aborting backup
- [PLAT-2101] fix: Save button doesn't throw field errors when all the fields are empty
- [PLAT-2102] Stop creation of KMS configs with duplicate name
- [PLAT-2103] [UI] [Encr at rest] AWS credentials enabled when IAM is checked in
- [PLAT-2109] Skip hostname validation in certificate
- [PLAT-2121] [UI] [KMS] UI displaying Successfully added even when it fails to add KMS config
- [PLAT-2133] Need to allow platform to repair a server that's stuck in a OOM state
- [PLAT-2148] Unable to change availability zone while configuring new universe
- [PLAT-2158] Getting rid of useHostnames
- [PLAT-2177] Avoid adding duplicate delete backup tasks to tasks queue
- [PLAT-2184] Fix a bug where new leader thinks of itself as remote instance
- [PLAT-2189] Fix universe creation on airgap install
- [PLAT-2214] fix: Cloud provider section overlapping with universe details
- [PLAT-2218] Removing and adding a single AZ to univ placement triggers a full move
- [PLAT-2241] Replace non-unicode symbols on traceback output decode
- [PLAT-2243] Fix chrony for airgap install (aws, gcp)
- [PLAT-2246] Fix Timezone UI Bugs
- [PLAT-2250] [Alerts] [UI] Alert Creation page is in loading state in brand new portal
- [PLAT-2251] Make health check run and store results to DB without health config
- [PLAT-2253] Skip delete release from local filesystem step in case of non-local release
- [PLAT-2257] Provision in AnsibleSetupServer cannot be retried if it fails in the middle because the default SSH port is no longer reachable.
- [PLAT-2259] [xcluster] [UI] Master Node Address unavailable on replication tab
- [PLAT-2265] Resolve missing list key and clean up YBMultiSelect styling.
- [PLAT-2269] Hide quick filter
- [PLAT-2270] Non-transactional backup with list of tables in YCQL fails
- [PLAT-2274] [UI] No way to remove region from provider
- [PLAT-2294] [UI] Disable toggle TLS / cert rotation dialogs in UI for k8s universes
- [PLAT-2298] Allow outOfOrder migrations and ignoreMissingMigrations
- [PLAT-2319] Remove sensitive data from platform log while during universe creation using s3 package
- [PLAT-2320] Fix Prometheus links to have correct timestamp
- [PLAT-2321] [PLAT-2314] Fix backup util behavior on error for onprem providers
- [PLAT-2323] Preparation subtask with error message should be added only if no other subtasks exist
- [PLAT-2325] Fix broken migration in older releases
- [PLAT-2326] Incorrect masters selection for GP universe
- [PLAT-2328] Reuse instance in cloud provider if it matches the universe ID and node ID tags.
- [PLAT-2344] Fix preflight checks
- [PLAT-2349] [UI] Blank screen when trying to add read replica to an universe
- [PLAT-2351] Disable upgrade to systemd for manually provisioned universes
- [PLAT-2373] Get rid of repeated migrations
- [PLAT-2376] Fixing resize node to utilize ssh_user param
- [PLAT-2400] Universe scaling out by adding nodes keeps new nodes in blacklisted state.
- [PLAT-2434] Fix omitted log configurations after application start
- [PLAT-2514] Fix leaderless and underreplicated tablet alerts in case master quorum is missing
- [PLAT-2348] Fix UT failure during change in max size of config name
- [T9574] [T8510] Force delete backup Configuration after deleting backups and schedule associated with it
Database
- [8660] [DocDB] Tablet splitting thresholds should be determined based on number of nodes in a table's placement policy
- [9665] [YSQL] ysql_dump should not need master addresses to be passed explicitly.
- [9929] [2809] [2742] [YSQL] Fix explicit row locking semantics + support SKIP LOCKED for REPEATABLE READ isolation level
- [10163] [DST] Disable large transaction batch apply until it does not affect tablet splitting
- [10193] [YSQL] Disallow deleting tablespace used by tablegroups
- [10236] [YSQL] Properly assign junk_filter_needed value based on type of Relation
- [10256] [DST] [PITR] Fix Incorrect Restore logic for System Catalog tables
- [10262] [DocDB] Tablespace task in the YB-Master must first check whether ysql is enabled
- [10303] [YSQL] fix unexpected transaction state error
- [10472] Use curl instead of wget for client download script
- [10519] Refresh master leader after election
- [10519] Reset master leader on meta cache timeouts
- [10521] [YSQL] Always republish migrations
- [10591] [YSQL] COPY for partitioned table writes to parent instead of child nodes
- [10622] [DocDB] Fix ASAN build
- [10629] [YSQL] Ensure follower reads go to the follower for reads from the index
- [10631] [YSQL] Hide index backfill auth key in logs
- [10646] [YSQL] fix client backfill timeout
- [10737] [backup] [YSQL] yb_backup must use '--masters' argument for ysql_dump.
- [10741] [YSQL] change ybgin substitution msg to LOG
- [10769] [Geo] Fixed data race when updating transaction status hash
- [10185] Check for schema column id mismatch in SetupUniverseReplication
Known Issues
Yugabyte Platform
-
[PLAT-2487] When you add read replicas, the instance type and volume may not show any values. Re-select the provider and the values should populate correctly as primary cluster.
-
[PLAT-2490] You may see a "Warning: cronjobs are not active on some nodes" message on the Universe health check page, even when cron jobs are active on the database nodes. To remove the warning:
-
For Replicated- and Yugabundle-based Platform installs, launch the
devops/bin/edit_universe_details.py
script and globally replace the value ofcronsActive
field fromtrue
tofalse
. The number ofcronsActive
entries corresponds to the number of database nodes in your universe — you need to replace all of them. -
For Kubernetes-based Platform installs, please contact Yugabyte support for help as this requires modifying the values in the database directly.
-
Database
N/A