yugabyte/yugabyte-db v2.9.1

v2.9.1.0 (Released Oct 29, 2021)

on GitHub

Note: To upgrade a pre-version 2.9.0.0 Yugabyte Platform or universe instance deployed on Kubernetes that did not specify a storage class override, you need to override the storage class Helm chart value (which is now "", the empty string) and set it to the previous value, "standard".

For Yugabyte Platform, the class is yugaware.storageClass. For YugabyteDB, the classes are storage.master.storageClass and storage.tserver.storageClass.

v2.9.1.0 - Oct 29, 2021

Build: 2.9.1.0-b140

Downloads

Docker

docker pull yugabytedb/yugabyte:2.9.1.0-b140

New Features

Yugabyte Platform

• [9124] Create REST endpoints to manage async replication relationships
• [9733] [Alerts] Implement alert listing
• [PLAT-417] Add support for Ubuntu 18.04 distributions
• [PLAT-1573] Adding 'create new cert' in enable TLS new feature
• [PLAT-1620] Added secondary subnet for allowing two network interfaces
• [PLAT-1669] Initial OEL 8 support
• [PLAT-1695] Create new API endpoint to be able to query logs by Universe
• [PLAT-1753] Enable taking backups using custom ports

Database

• [1127] [YSQL] Collation Support (part 2)
• [1127] [YSQL] Collation Support (part 3)
• [6541] [YSQL] Enable row-locking feature in CURSOR
• [7850] [YSQL] create new access method ybgin
• [8402] [YSQL] change gin to ybgin for YB indexes
• [YSQL] Foreign Data Wrapper Support

Improvements

Yugabyte Platform

• [5236] [PLAT-59] Allow log levels to be changed through POST /logging_config endpoint
• [7396] Splitting up create/provision tasks to delete orphaned resources
• [7645] [PLAT-523] Show error summary at the top of the health check email
• [9131] Enable/disable YCQL endpoint while universe creation and force password requirement
• [9407] [PLAT-386] Implement base YSQL/YCQL alerts
• [9580] Add restore_time field for all universes.
• [9613] Update UI to accommodate Authentication changes
• [9668] Alert configurations implement missing parts and few small changes
• [9794] [PLAT-1530] Creates static IP during cluster creation for cloud free tier clusters. Releases IPs on deletion.
• [9892] Mask sensitive gflag info
• [9978] Platform UI: Change stop backup icon and label to abort icon and label.
• [CLOUDGA-2345] Implement MDC propagation and add request/universe ID to MDC
• [PLAT-525] Add IP address to SAN of node certificates
• [PLAT-541] Allow configuring no destination for alert config + UI improvements
• [PLAT-1523] Update Alert APIs to be consistent with UI terminology
• [PLAT-1528] Change YWError handler to default to JSON response on client error.
• [PLAT-1546] [PLAT-1547] [PLAT-1571] [PLAT-1572] Add API docs for UniverseClustersController, and other misc fixes
• [PLAT-1549] Add (non-generated client, "simple") Python API examples
• [PLAT-1549] Cleanup list/create provider API usage examples
• [PLAT-1555] Add Python API client example for create Universe
• [PLAT-1555] Add Python API client example for list Universe
• [PLAT-1556] List Storage Configs Create Scheduled backup examples
• [PLAT-1582] [Alert] Limit Severity to maximum 2(Severe/warn), now we can add multiple severity's but after edit we are displaying only 2 (1 Severe/1 Warn)
• [PLAT-1611] Add python depedencies required for executing external scripts
• [PLAT-1647] Provide more details for default channel on UI
• [PLAT-1664] Enable new alert UIs and remove deprecated alert UI + configs from Health tab + config from replication tab
• [PLAT-1691] Set oshi LinuxFileSystem log level to ERROR
• [PLAT-1691] Task, API and thread pool metrics
• [PLAT-1705] Add auditing and transaction for /register API action
• [PLAT-1723] Allow disabling prometheus management + write alerts and metrics effectively
• [PLAT-1766] [Alerts] UI: Cleanup
• [PLAT-1774] Add a customer ID field in Customer Profile page
• [PLAT-1791] Use hibernate validator for all alert related entities
• [PLAT-1818] Add pagination to Tables tab and add classNames
• Added new AWS regions to metadata files.
• Hooking GCP Create Method into Create Root Volumes method

Database

• [2220] [YSQL] Enabling relation size estimation for temporary tables in optimizer
• [5492] yb-admin: Added error message when attempting to create snapshot of YCQL system tables
• [7612] [DocDB] Allow TTL-expired SST files that are too large for compaction to be directly expired
• [7612] [DocDB] Modified compaction file filter to filter files out of order
• [7889] Reduce timeout for ysql backfill.
• [8043] [YBase] Remove information about LB skipping deleted tables from the admin UI
• [8162] [YSQL] Support single-request optimization for UPDATE with RETURNING clause
• [8229] [backup] repartition table if needed on YSQL restore
• [8452] Speed up restoring YSQL system catalog
• [8501] [DocDB] Add metric to monitor server uptime
• [8508] [DocDB] moved GetSplitKey from TabletServerAdminService into TabletServerService
• [8580] [9489] [YSQL] Inherit default PGSQL proxy bind address from rpc bind address
• [8804] [YSQL] [backup] Support in backups the same table name across different schemas.
• [8807] [YBase] Add a limit on number of metrics for the prometheus metrics endpoint
• [8979] [DocDB] Improve master load balancer state presentation
• [9279] [YSQL] Enable -Wextra on pgwrapper
• [9279] [YSQL] Enable -Wextra on yql folder
• [9418] [xCluster] Add cdc_state Schema Caching to Producer Cluster
• [9439] [YBase] Allow sst-dump to decode DocDB keys and dump data in human readable format
• [9467] [YSQL] Increase scope of cases where transparent retries are performed
• [9685] [xCluster] Make delete_universe_replication fault tolerant
• [9739] Added placement info to /api/v1/tablet-servers
• [9746] Set WAL footer close_timestamp_micros on Bootstrap
• [9762] [Part-1] Populate partial index predicate in "options" column of system_schema.indexes
• [9803] [YSQL] Import Avoid trying to lock OLD/NEW in a rule with FOR UPDATE.
• [9831] [YSQL] Import Fix broken snapshot handling in parallel workers.
• [9862] [PITR] Allow consecutive restore
• [9862] Allow PITR in conjunction with tablet split
• [9899] [YSQL] Import Fix corner-case uninitialized-variable issues in plpgsql.
• [9911] [YSQL] Import In pg_dump, avoid doing per-table queries for RLS policies.
• [9922] [YSQL] Import Fix float4/float8 hash functions to produce uniform results for NaNs.
• [9926] [YSQL] Import Disallow creating an ICU collation if the DB encoding won't support it.
• [9935] [YSQL] Import Fix bitmap AND/OR scans on the inside of a nestloop partition-wise join.
• [9936] Remove YBClient from Postgres: Introduce PgClient and implement ReserveOids using it; Open table via PgClient; Remove all direct YBClient usage from PgSession
• [9966] [YSQL] Import Rearrange pgstat_bestart() to avoid failures within its critical section.
• [9995] [YSQL] Import Fix EXIT out of outermost block in plpgsql.
• [10025] [YSQL] Import jit: Do not try to shut down LLVM state in case of LLVM triggered errors.
• [10034] [YSQL] Preserve operation buffering state in case of transparent retries
• [10064] [xCluster] Lag Metric Improvements
• [10111] [YSQL] Import Force NO SCROLL for plpgsql's implicit cursors.
• [10121] [YSQL] Import Avoid misbehavior when persisting a non-stable cursor.
• [YSQL] Import Fix performance bug in regexp's citerdissect/creviterdissect.

Bug Fixes

Yugabyte Platform

• [1525] New Universe creation gets public IP assigned even with flag = false
• [1598] [UI] Suggested Default File Path for CA Signed Certificate and Private Key is Incorrect
• [7573] [PLAT-482] Health Checks should run when Backup/Restore Tasks are in progress
• [7738] [PLAT-611] Health checks can overlap with universe update operations started after them
• [8510] Allow the deletion of Failed Backups
• [9014] [PLAT-509] Refresh Pricing data for Azure provider seems to be stuck
• [9315] [PLAT-521] BackupsController: small fixes required
• [9366] [PLAT-368] Disable Delete Configuration button for backups when in use.
• [9850] [YW] Correct the node path (#9864)
• [CLOUDGA-1893] fix client-to-node cert path in health checks
• [PLAT-253] Fix the backupTable params while creating Table backups using Apis.
• [PLAT-253] Fix universe's backupInprogress flag to avoid multiple backup at a time due to low frequency scheduler.
• [PLAT-289] Stopped node should not allow Release action
• [PLAT-580] Fix create xCluster config API call
• [PLAT-599] Fix error messages in alert destination and configuration services
• [PLAT-1520] Stop displaying external script schedule among Backup Schedules.
• [PLAT-1522] Fix s3 release breakage
• [PLAT-1549] [PLAT-1697] Fix Stop backup race condition. Add non-schedlued backup examples
• [PLAT-1559] Stop the external script scheduler if the universe is not present.
• [PLAT-1563] Fix instance down alerts + make sure instance restart alert is not fired on universe operations
• [PLAT-1578] Do not specify storage class (use default if provided)
• [PLAT-1586] [Alert] Able to add multiple alert configuration with same name. Add duplicate check for alert configuration name
• [PLAT-1599] [UI] Root Certificate and node-node and client-node TLS missing on Edit Universe
• [PLAT-1603] [Platform]YBFormInput's OnBlur throws error on AddCertificateForm
• [PLAT-1605] Fix duplicate alert definitions handling + all locks to avoid duplicates creation
• [PLAT-1606] Disk name too long for Google Cloud clone disk
• [PLAT-1613] Alerts: Logs filled with NPE related to "Error while sending notification for alert "
• [PLAT-1617] Added GCP region metadata for missing regions.
• [PLAT-1617] Fix issue with GCP Subnet CIDR
• [PLAT-1619] Check for FAILED status in wait_for_snapshot method.
• [PLAT-1621] Health check failed in K8s portal
• [PLAT-1625] Fix task details NPE
• [PLAT-1626] Skip preprovision for systemd upgrade.
• [PLAT-1631] [Alert] Universe filter is not working in Alert Listing
• [PLAT-1638] Fix naming convention for external script endpoints as per our standards
• [PLAT-1639] [PLAT-1681] Make proxy requests async to keep them from blocking other requests
• [PLAT-1639] [PLAT-1681] Reduce log spew from akka-http-core for proxy requests.
• [PLAT-1644] Fix k8s universe creation failure for platform configured with HA
• [PLAT-1646] Remove Unsupported Instance types from pull down menu for Azure
• [PLAT-1650] Added yum lock_timeout to prevent yum lockfile errors for use_custom_ssh_port.yml
• [PLAT-1653] Fix region get/list.
• [PLAT-1656] [UI] [Alert] Group Type filter is not working in Alert Listing
• [PLAT-1661] Fix alert messages for notification failures
• [PLAT-1667] Platform should not scrape all per-table metrics from db hosts (part 2)
• [PLAT-1668] Yugabundle failing because can't find YWErrorHandler
• [PLAT-1682] Fix node comparison function from accessing undefined cluster
• [PLAT-1687] ALERT: Not able to create destination channel using "default recipients + default smtp settings + empty email field"
• [PLAT-1694] Fix Intermittent failure to back up k8s universe
• [PLAT-1707] Fix performance issue
• [PLAT-1715] Check for YB version only for 2.6+ release DB
• [PLAT-1717] Full move fails midway if system tablet takes more than 2 mins to bootstrap
• [PLAT-1721] Stop storage type from automatically changing when instance type is changed
• [PLAT-1726] Allow user to completely remove all gFlags after addtion of several gFlags.
• [PLAT-1730] Fix resize node logic for RF1 clusters
• [PLAT-1736] Create default alert configs and destination on DB seed
• [PLAT-1737] "This field is required" error message is shown on alert configuration creation with default threshold == 0
• [PLAT-1746] Delete prometheus_snapshot directory once platform backup package is created
• [PLAT-1757] Health Check failure message has Actual and expected values interchanged
• [PLAT-1761] Fix alert message in case of unprovisioned nodes
• [PLAT-1768] Universe tasks take lot more time because thread pool executors do not reach max_threads
• [PLAT-1780] Redact YSQL/YCQL passwords from task_info table.
• [PLAT-1793] DB Error logs alert
• [PLAT-1796] Edit Universe page has password fields editable
• [PLAT-1802] Replication graphs stopped showing on replication tab (replicated.yml change)
• [PLAT-1804] Fix 'Querying for {} metric keys - may affect performance' log
• [PLAT-1816] Forward port restricted user creation to master
• [PLAT-1819] [PLAT-1828] Release backup lock when Platform restarts, and update Backup state
• [PLAT-1829] [ycql/ysql] auth password: wrong error message
• [PLAT-1833] Fix missing create time on alert configuration creation issue
• [PLAT-1839] Fix typo in DB migration
• [PLAT-1892] Make error alert be disabled by default
• [PLAT-1969] [UI] Universe creation - Create button is disabled when YSQL/YCQL auth is disabled
• Backup and Restore failing in k8s auth enabled environment
• Fix NPE in VM image upgrade for TLS enabled universes
• Use TaskInfo instead of CustomerTask in shouldIncrementVersion check

Database

• [1252] Do not link with system libpq
• [5920] Fix bootstrapping with preallocated log segment
• [7528] [YSQL] Error out when Tablespaces are set for colocated tables
• [8675] [DocDB] Prevent tablet splitting when there is post split data
• [8772] Fix fatal that occurs when running alter_universe_replication and producer master has
• [9061] [DocDB] Master task tracking should point to the table it is operating on
• [9216] [YSQL] Fix NULL pointer access in case of failed test
• [9436] [YSQL] Statement reads rows it has inserted
• [9475] Fetch Universe Key From Masters on TS Init
• [9616] Fix master crash when restoring snapshot schedule with deleted namespace
• [9655] [xCluster] Label cdc streams with relevant metadata
• [9743] Fix universe reset config option (#9863)
• [9781] Mark snapshot as deleted if tablet was removed
• [9782] [DocDB] Tablet Splitting - Wait for all peers to finish compacting during throttling
• [9786] Universe Actions -> Add Read Replica is failing
• [9789] [DocDB] Load Balancer should use tablet count while looking tablets to move
• [9802] [xCluster] Set proper deadline for YBSession in CDCServiceImpl
• [9806] DocDB: fixed Batcher::FlushBuffersIsReady
• [9812] [YSQL] Check database is colocated before adding colocated option for Alter Table
• [9822] DocDB: Check table pointer is not nullptr before dereferencing
• [9855] [DocDB] Set aborted subtransaction data on local apply
• [9860] [YSQL] fix limit vars to uint64
• [9865] Fix internal retry of kReadRestart for SELECT func() with a DML in the func
• [9867] [YSQL] Fix double type overflow in case of SET yb_transaction_priority_lower_bound/yb_transaction_priority_upperr_bound command
• [9906] [YSQL] Fix not being able to add a range primary key
• [9909] [YSQL] further fix backup restore for NULL col attr
• [9924] [YSQL] always check schema name on backup import
• [9927] YCQL - Handle unset correctly
• [9932] [YSQL] Initialize t_ybctid field in acquire_sample_rows()
• [9933] [Part-0] Update logic for using num_tablets from internal or user requests.
• [9933] [YCQL] [Part-1] DESC TABLE does not directly match the "CREATE TABLE" command for number of tablets.
• [9934] [DocDB] Don't update rocksdb_dir on Remote Bootstrap
• [9936] Fix ysql_dump in encrypted k8s environment
• [9936] Fix ysql_dump in TLS encrypted environment
• [9940] DocDB: use correct kv_store_id for post-split tablets
• [9947] [YSQL] remove runtime tag for ysql_disable_index_backfill
• [9963] backup: fix to reallow YEDIS on restore
• [9965] [YSQL] Fix copy/paste error causing incorrect conversion
• [9981] Fix transaction coordinator returning wrong status hybrid time
• [10042] [backup] allow system table for YEDIS restore
• [10051] DocDB: use RETURN_NOT_OK on an unchecked status
• [10085] YSQL fix FATAL caused by wrong sum pushdown
• [10098] [YSQL] Fix index creation on temp table via ALTER TABLE
• [10139] ybase: Avoid unnecessary table locking in CatalogManager::DeleteYsqlDBTables
• [ybase] Properly pass number of tables via MetricPrometheusOptions
• [YSQL] [9572] Correctly determine is_yb_relation for row-marked relations when preparing target list
• Fix for resource leaks
• Fixed bug in yb-ctl for stopping processes, when os.kill raises an exception
• Make SSH wait behavior consistent across operations

Known Issues

Yugabyte Platform

N/A

Database

N/A