diff --git a/CHANGELOG.md b/CHANGELOG.md
index 8635540c2..a708d0e7a 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -16,10 +16,24 @@ Changes
 
 .. towncrier release notes start
 
+21.03.1 (2021-03-31)
+--------------------
+
+### Fixes
+* Fix a missing reference fix for renaming of `gateway` to `manager.api` ([#409](https://github.com/lablup/backend.ai-manager/issues/409))
+* Refactor the manager CLI initialization steps and promote `generate-keypair` as a regular `mgr` subcommand ([#411](https://github.com/lablup/backend.ai-manager/issues/411))
+* Fix an internal API mismatch for our SQLAlchemy custom enum types ([#412](https://github.com/lablup/backend.ai-manager/issues/412))
+* Fix a regression in session cancellation and kernel status updates after SQLAlchemy v1.4 upgrade ([#413](https://github.com/lablup/backend.ai-manager/issues/413))
+
+### Miscellaneous
+* Fix the examples for the storage proxy URL configurations in the `manager.config` module ([#410](https://github.com/lablup/backend.ai-manager/issues/410))
+* Update sample configurations for etcd ([#414](https://github.com/lablup/backend.ai-manager/issues/414))
+
+
 21.03.0 (2021-03-29)
 --------------------
 
-The v21.03.0 release is an integrated update for all features and fixes already applied to v20.09 series as additional patches, except that it now runs on top of Python 3.9.
+The v21.03.0 release is an integrated update for [all features and fixes already applied to v20.09 series](https://github.com/lablup/backend.ai-manager/blob/20.09/CHANGELOG.md) as additional patches, except that it now runs on top of Python 3.9.
 
 ### Features
 * Add `usage` field to the `StorageVolume` graph object schema to provide the total capacity and usage ([#389](https://github.com/lablup/backend.ai-manager/issues/389))
@@ -49,337 +63,3 @@ The v21.03.0 release is an integrated update for all features and fixes already
   - There is no changes in the TOML configuration but the `BACKEND_GATEWAY_NPROC` environment variable is renamed to `BACKEND_MANAGER_NPROC`.
   - You must update the license activator plugin (only the Enterprise edition users).
   - You may need to update SSO/webapp plugins to adapt with the new import paths. ([#408](https://github.com/lablup/backend.ai-manager/issues/408))
-
-
-20.09.4 (2021-01-08)
---------------------
-
-### Features
-* Add more indexes to reduce the latency for `kernels.cluster_role` involving queries ([#385](https://github.com/lablup/backend.ai-manager/issues/385))
-
-
-20.09.3 (2021-01-04)
---------------------
-
-### Fixes
-* Fix reloading of new container registry configs when rescanning, which is a regression due to registry abstraction for Harbor v2 support ([#382](https://github.com/lablup/backend.ai-manager/issues/382))
-* Fix `cluster_idx` for single-node sessions with only one container (`main0` should be `main1` to be consistent with multi-node sessions) ([#383](https://github.com/lablup/backend.ai-manager/issues/383))
-
-
-20.09.2 (2020-12-30)
---------------------
-
-### Fixes
-* Fix the duplicate passed/failed predicate list in `status_data` ([#379](https://github.com/lablup/backend.ai-manager/issues/379))
-* Fix a missing kernel ID field during cancel-termination upon failed creation of multi-container sessions ([#380](https://github.com/lablup/backend.ai-manager/issues/380))
-
-
-20.09.1 (2020-12-28)
---------------------
-
-### Fixes
-* Put aiodocker in the default dependency, not testing. ([#377](https://github.com/lablup/backend.ai-manager/issues/377))
-* Remove use of deprecated asyncio API (`asyncio.Task.{all_tasks,current_task}` -> `asyncio.{all_tasks,current_task}`) to run with Python 3.9 ([#378](https://github.com/lablup/backend.ai-manager/issues/378))
-
-
-20.09.0 (2020-12-27)
---------------------
-
-### Features
-* Implement `max_container_per_session` keypair resource policy as a part of predicate checks ([#376](https://github.com/lablup/backend.ai-manager/issues/376))
-
-### Fixes
-* Additional stabilization of multi-node multi-container support ([#376](https://github.com/lablup/backend.ai-manager/issues/376))
-  - Fix destruction of multi-container sessions when using multiple nodes
-  - Adopt `MultiAgentError` to represent and handle partial failures of multi-container session creation
-
-
-20.09.0rc3 (2020-12-26)
------------------------
-
-### Fixes
-* Stabilize multi-container sessions further. ([#373](https://github.com/lablup/backend.ai-manager/issues/373))
-  - Include predicate check results in `kernels.status_data` for better user-side diagnosis
-  - Fix further race condition in the session-level creation routines, by adopting `session_creation_id` like the `kernel_creation_id` introduced in #374.
-  - Remove the no longer used `agents.clusterized` column as it is no longer included in agent heartbeats after k8s branch refactoring.
-  - Use dual DB connections to rollback agent-specific (resource) updates but to keep kernel-specific (status) updates when scheduling failures occur.
-  - Fix a potential duplicate of DB cursors because use-after-free of a DB connection in the manager scheduler.
-  - Relax row-level locks in scheduler in favor of the `REPEATABLE READ` isolation level.
-  - Synchronize the candidate agent list when iterating over multiple kernels for a session in the multi-node mode, because they should now be updated
-    according to the transaction savepoints.
-  - Fix blocked proceeding to the next scaling group when a scheduling failure occurs in a scaling group.
-  - Shield the DB queries in the events queue handlers and now interrupting the manager service works better.
-
-
-20.09.0rc2 (2020-12-24)
------------------------
-
-### Fixes
-* Fix races of kernel creation events by attaching a unique creation request ID to distinguish and catch the events by the caller manager instance ([#374](https://github.com/lablup/backend.ai-manager/issues/374))
-
-
-20.09.0rc1 (2020-12-23)
------------------------
-
-### Fixes
-* Unable to request watcher API due to incorrect reference to watcher endpoint. ([#370](https://github.com/lablup/backend.ai-manager/issues/370))
-* Replace deprecated reference of kernels.c.role with kernels.c.cluster_role in legacy compute session list query. ([#371](https://github.com/lablup/backend.ai-manager/issues/371))
-* Fix various bugs related to multi-node execution paths and provide richer error information with agent-side errors as a new JSON column `status_data` in the kernels table ([#372](https://github.com/lablup/backend.ai-manager/issues/372))
-
-
-20.09.0b6 (2020-12-21)
-----------------------
-
-### Features
-* Add a new API path for per-agent hardware metadata queries ([#366](https://github.com/lablup/backend.ai-manager/issues/366))
-
-
-20.09.0b5 (2020-12-20)
-----------------------
-
-### Features
-* Support destroying PREPARING/TERMINATING/ERROR sessions when forced parameter is delivered as True. ([#363](https://github.com/lablup/backend.ai-manager/issues/363))
-
-### Fixes
-* Fix a bug that prevents purge users and/or groups due to access to vfroot filesystem from manager directly, instead of delegating the tasks to storage-proxy. ([#361](https://github.com/lablup/backend.ai-manager/issues/361))
-* Fix a race condition that sometimes corrupts the kernel status to be stuck at PREPARING even when kernels are successfully created ([#368](https://github.com/lablup/backend.ai-manager/issues/368))
-
-
-20.09.0b4 (2020-12-19)
-----------------------
-
-### Fixes
-* Fix a regression in image list graph queries due to a missing renaming of local_config ([#367](https://github.com/lablup/backend.ai-manager/issues/367))
-
-
-20.09.0b3 (2020-12-18)
-----------------------
-
-### Fixes
-* Update aiotools to v1.1.1 to fix a potential memory leak in long-lived taskgroup instances ([#362](https://github.com/lablup/backend.ai-manager/issues/362))
-* Fix a hang-up issue when shutting down the gateway daemon with running service-port streaming sessions ([#365](https://github.com/lablup/backend.ai-manager/issues/365))
-
-### Miscellaneous
-* Update dependencies and adapt with the new pip resolver ([#359](https://github.com/lablup/backend.ai-manager/issues/359))
-* Migrate to GitHub Actions completely for all CI jobs including tests and PyPI deployment ([#364](https://github.com/lablup/backend.ai-manager/issues/364))
-
-
-20.09.0b2 (2020-12-07)
-----------------------
-
-### Features
-* Add support for Harbor v2 and generalize internal abstractions of container registries, making it easier to add new registries in the future ([#357](https://github.com/lablup/backend.ai-manager/issues/357))
-
-### Fixes
-* Update fixtures to put the default registry as `cr.backend.ai` and remove unused ones ([#358](https://github.com/lablup/backend.ai-manager/issues/358))
-
-
-20.09.0b1 (2020-12-02)
-----------------------
-
-### Fixes
-* Improve handling of storage-manager invocation failures using a new `VFolderOperationFailed` API exception. ([#352](https://github.com/lablup/backend.ai-manager/issues/352))
-* Fix missing replacement of `config_server` to `shared_config` in GraphQL context objects and references via `request.app['registry'].config_server` in some API handlers ([#354](https://github.com/lablup/backend.ai-manager/issues/354))
-* Fix a regression of fetching `live_stat` GraphQL field values ([#355](https://github.com/lablup/backend.ai-manager/issues/355))
-* Improve the error message for image not found errors when creating new sessions so that users could guess the reason quicklier ([#356](https://github.com/lablup/backend.ai-manager/issues/356))
-
-
-20.09.0a4 (2020-11-16)
-----------------------
-
-### Features
-* New idle checker to automatically kill sessions using various criteria ([#341](https://github.com/lablup/backend.ai-manager/issues/341))
-
-### Fixes
-* Implement and use a new global timer which works regardless of the number of manager instances to sustain periodic tasks such as session scheduling and idle checks ([#341](https://github.com/lablup/backend.ai-manager/issues/341))
-* Replace legacy sess_id to session_name. ([#348](https://github.com/lablup/backend.ai-manager/issues/348))
-* Fix intermittent resource warnings from aiopg by replacing all `fetchone()` with `first()` and `scalar()` that closes the database cursor automatically always ([#349](https://github.com/lablup/backend.ai-manager/issues/349))
-* Update compute_plugins info from heartbeat for an ALIVE agent. ([#350](https://github.com/lablup/backend.ai-manager/issues/350))
-* Fix broken signout. ([#353](https://github.com/lablup/backend.ai-manager/issues/353))
-
-
-20.09.0a3 (2020-11-03)
-----------------------
-
-### Fixes
-* Fix a regression in container-logs API since session name/ID matching generalization ([#346](https://github.com/lablup/backend.ai-manager/issues/346))
-* Improve error handling in `gateway.wsproxy.TCPProxy` and use only aiohttp-supported WebSocket close codes to prevent client-side break-up ([#347](https://github.com/lablup/backend.ai-manager/issues/347))
-
-
-20.09.0a2 (2020-10-30)
-----------------------
-
-### Features
-* Add a new API `/session/{session_ref}/shutdown-service` for shutting down running in-container services ([#327](https://github.com/lablup/backend.ai-manager/issues/327))
-* Add hooking point for AUTHORIZE with FIRST_COMPLETED requirement. ([#339](https://github.com/lablup/backend.ai-manager/issues/339))
-
-### Fixes
-* Unable to invite user to a vfolder, which is shared with another user. ([#340](https://github.com/lablup/backend.ai-manager/issues/340))
-* Fix a regression of the manager status API due to internal etcd data structure changes, by returning the first-seen manager ID (for HA scenarios) to keep the compatibility of API semantics ([#342](https://github.com/lablup/backend.ai-manager/issues/342))
-* Fix a regression in paginated list GraphQL query for vfolders due to missing code updates ([#343](https://github.com/lablup/backend.ai-manager/issues/343))
-* Ensure removal of vfolders from the database when storage-proxy returns an error (e.g., not found) ([#344](https://github.com/lablup/backend.ai-manager/issues/344))
-* Update dependencies (aiohttp~=3.7.0, trafaret~=2.1). ([#345](https://github.com/lablup/backend.ai-manager/issues/345))
-
-
-20.09.0a1 (2020-10-06)
-----------------------
-
-### Breaking Changes
-* The latest API version is now bumped to `v6.20200815`. ([#0](https://github.com/lablup/backend.ai-manager/issues/0))
-* Configuration/DB changes are required for storage proxies ([#312](https://github.com/lablup/backend.ai-manager/issues/312))
-  - To use vfolders, now the storage proxy must be installed and configured.
-  - The "volumes" key in the etcd must include storage proxy configurations,
-    as demonstrated in the `config/sample.volume.json` file.
-  - All vfolder hosts are now in the format of "{proxy-name}:{volume-name}"
-    where proxy-name is the key in the etcd and volume-name values are retrieved from
-    the storage proxies at runtime.
-  - All `allowed_vfolder_hosts` configurations in the database must be updated to
-    the above new vfolder host format.
-  - Clients should use the same vfolder host format when making API requests.
-
-### Features
-* Add support for multi-container sessions ([#217](https://github.com/lablup/backend.ai-manager/issues/217))
-* Add a generic query filter expression language parser (`ai.backend.manager.models.minilang.queryfilter`) for GraphQL paginated list queries using [the Lark parser framework](https://github.com/lark-parser/lark) ([#305](https://github.com/lablup/backend.ai-manager/issues/305))
-* Add support for storage proxies ([#312](https://github.com/lablup/backend.ai-manager/issues/312), [#337](https://github.com/lablup/backend.ai-manager/issues/337))
-  - Storage proxies have a multi-backend architecture, so now storage-specific optimizations such as per-directory quota, performance measurements and fast metadata scanning all becomes available to Backend.AI users.
-  - Offload and unify vfolder upload/download operations to storage proxies via the HTTP ranged queries and the tus.io protocol.
-  - Support multiple storage proxies configured via etcd, and each storage proxy may provide multiple volumes in the mount points shared with agents.
-  - Now the manager instances don't have mount points for the storage volumes, and mount/fstab management APIs skip the manager-side queries and manipulations.
-* Include user information (full_name) in keypair gql query. ([#313](https://github.com/lablup/backend.ai-manager/issues/313))
-* Add an endpoint that allows users to leave a shared virtual folder ([#317](https://github.com/lablup/backend.ai-manager/issues/317))
-* Make the `mgr dbshell` command to be smarter to auto-detect the halfstack db container and use the container-provided psql command for maximum compatibility, with an optinal ability to set the container ID or name explicitly via `--psql-container` option and forward additional arguments to the psql command ([#318](https://github.com/lablup/backend.ai-manager/issues/318))
-* Script to migrate /vroot/local to ex. /vroot/vfs structure according with new Storage Proxy implementation. ([#319](https://github.com/lablup/backend.ai-manager/issues/319))
-* Make the maximum websocket message size configurable, which affects the operation of streaming APIs including service-port proxy ([#320](https://github.com/lablup/backend.ai-manager/issues/320))
-* Add a new vfolder API to clone a vfolder and a property to vfolders for specifying cloneable or not ([#323](https://github.com/lablup/backend.ai-manager/issues/323), [#338](https://github.com/lablup/backend.ai-manager/issues/338))
-* Add `quota` argument when creating vfolders (currently only supported in the xfs storage backend) ([#325](https://github.com/lablup/backend.ai-manager/issues/325))
-* Add support for listing/updating/deleting/creating domain dotfiles and group dotfiles ([#329](https://github.com/lablup/backend.ai-manager/issues/329))
-* Make vfolder's mkdir to accept and deliver parents and exist_ok option. ([#336](https://github.com/lablup/backend.ai-manager/issues/336))
-
-### Fixes
-* Prevent purging a user, group, and domain if they have active sessions and/or
-  their bound resources, such as virtual folders, are being used by other sessions. ([#306](https://github.com/lablup/backend.ai-manager/issues/306))
-* Fix conversion of string query params to int/bool ([#307](https://github.com/lablup/backend.ai-manager/issues/307))
-* Do not create invitation if target user is already shared the folder. ([#308](https://github.com/lablup/backend.ai-manager/issues/308))
-* Ignore missing tags in the Docker Hub when scanning the metadata for public Docker images, which happens after deleting images from the hub ([#309](https://github.com/lablup/backend.ai-manager/issues/309))
-* Fix a regression of the GQL query to fetch the information about a single compute session ([#311](https://github.com/lablup/backend.ai-manager/issues/311))
-* Return shared memory in preset list API. ([#314](https://github.com/lablup/backend.ai-manager/issues/314))
-* Fix a regression in the server-side error logs query API due to a typo in the DB column name ([#315](https://github.com/lablup/backend.ai-manager/issues/315))
-* Fix ErrorMonitor plugin instance to match the updates in plugin class of backend.ai-common ([#316](https://github.com/lablup/backend.ai-manager/issues/316))
-* Deduplicate explicitly raised internal server errors in the daemon logs ([#322](https://github.com/lablup/backend.ai-manager/issues/322))
-* Log detailed error messages when graphQl exceptions are raised. ([#324](https://github.com/lablup/backend.ai-manager/issues/324))
-* Fix a regression of statistics API due to a wrong default value type ([#326](https://github.com/lablup/backend.ai-manager/issues/326))
-* Fix vfolder mounts for compute sessions using agent host paths provided by the storage proxy ([#328](https://github.com/lablup/backend.ai-manager/issues/328))
-* Hand over recursive parameter to storage proxy when deleting files. ([#330](https://github.com/lablup/backend.ai-manager/issues/330))
-* Validate service port declarations before starting the session ([#333](https://github.com/lablup/backend.ai-manager/issues/333))
-* Improve pickling of wrapped HTTP exceptions (`BackendError` and its derivatives) ([#334](https://github.com/lablup/backend.ai-manager/issues/334))
-* Show the container name when using the `mgr dbshell` cli command ([#335](https://github.com/lablup/backend.ai-manager/issues/335))
-
-
-20.03.0 (2020-07-28)
---------------------
-
-### Features
-* Add purge mutations for users, groups, and domains. ([#302](https://github.com/lablup/backend.ai-manager/issues/302))
-
-### Fixes
-* Raise explicit exception when no keypair is found during login. ([#297](https://github.com/lablup/backend.ai-manager/issues/297))
-* Adds context for the call to from_row in simple_db_mutate_returning_item. ([#300](https://github.com/lablup/backend.ai-manager/issues/300))
-* The session creation API now returns the session UUID as part of its response, in addition to the meaningless client-provided session alias (name) ([#303](https://github.com/lablup/backend.ai-manager/issues/303))
-
-### Miscellaneous
-* Return explicit message if user needs verification during login. ([#301](https://github.com/lablup/backend.ai-manager/issues/301))
-
-
-20.03.0rc1 (2020-07-23)
------------------------
-
-### Breaking Changes
-* Replace user's boolean-based `is_active` field to a enum-based `status` field to represent multiple user statuses with lifecycle state transitions, changing the signup API and related DB columns.
-* Change the argument type of `user_id` in `user_from_uuid` GraphQL query to the generic `ID` ([#299](https://github.com/lablup/backend.ai-manager/issues/299))
-
-### Fixes
-* Fix misuse of a legacy API in the `etcd move-subtree` manager command, which has unintentionally flattened nested dicts when inserting the data to the new location ([#295](https://github.com/lablup/backend.ai-manager/issues/295))
-* Mask sensitive keys when logging API parameters for debugging ([#296](https://github.com/lablup/backend.ai-manager/issues/296))
-* A set of API behavior fixes ([#299](https://github.com/lablup/backend.ai-manager/issues/299))
-  - Fix missing conversion of `user_id` argument to UUID type when resolving the `user_from_uuid` GraphQL query
-  - Exclude cancelled sessions from the active session count in the manager status API
-  - Fix missing update for user's `status_info` as "admin-requested" when marked as deleted by administrators
-
-### Miscellaneous
-* Refactor session usage api. ([#298](https://github.com/lablup/backend.ai-manager/issues/298))
-
-
-20.03.0b2 (2020-07-02)
-----------------------
-
-### Breaking Changes
-* Apply the plugin API v2 -- all stats/error/webapp/hook plugins must be updated along with the manager ([#291](https://github.com/lablup/backend.ai-manager/issues/291))
-
-### Features
-* Group-shared vfolders named ".local" (while other dot-prefixed names are not allowed for group vfolders) now have per-user subdirectories (using user UUIDs) for per-user persistent package installation reusable across different sessions ([#224](https://github.com/lablup/backend.ai-manager/issues/224))
-* Download a directory from vfolder as an archive. ([#278](https://github.com/lablup/backend.ai-manager/issues/278))
-* Execute POST_SIGNUP hook after sign up. ([#286](https://github.com/lablup/backend.ai-manager/issues/286))
-* Pass Accept-Language header to post-signup hook. ([#287](https://github.com/lablup/backend.ai-manager/issues/287))
-* Add a feature to kick off user who is shared virtual folder. ([#288](https://github.com/lablup/backend.ai-manager/issues/288))
-* Add a global announcement get/set API which stores an arbitrary string in the "manager/announcement" etcd key. ([#289](https://github.com/lablup/backend.ai-manager/issues/289))
-* Add option to reserve batch session. ([#290](https://github.com/lablup/backend.ai-manager/issues/290))
-* Add modified_at column to users and keypairs table. ([#293](https://github.com/lablup/backend.ai-manager/issues/293))
-* Add super-admin APIs to temporarily exclude specific agents from scheduling and include later by changing the `schedulable` column of the `agents` table. ([#294](https://github.com/lablup/backend.ai-manager/issues/294))
-
-### Fixes
-* Destroying sessions now returns periodically collected statistics instead of last-moment statistics ([#279](https://github.com/lablup/backend.ai-manager/issues/279))
-* Apply batching to statistics synchroniztion to minimize time used inside DB transactions by batching and splitting transaction blocks ([#280](https://github.com/lablup/backend.ai-manager/issues/280))
-* Add fallback if any keys under last_stat have None value during statistics calculation. ([#282](https://github.com/lablup/backend.ai-manager/issues/282))
-* Fix the responses for `compute_session` GQL query when there are no or multiple matching sessions ([#283](https://github.com/lablup/backend.ai-manager/issues/283))
-* ([#284](https://github.com/lablup/backend.ai-manager/issues/284))
-  - Fix too many scheduler invocation with stale periodic scheduling ticks after manager restarts/maintenance
-  - Split DB transactions for schedulers as fine-grained, small-sized transactions to cope with a large number of pending sessions
-* Ensure resolving of kernel IDs before doing complex operations in `AgentRegistry` to prevent potential races against the unique constraints about kernel statuses and user-defined session names under heavy loads ([#292](https://github.com/lablup/backend.ai-manager/issues/292))
-
-
-20.03.0b1 (2020-05-12)
-----------------------
-
-### Breaking Changes
-* Now it runs on Python 3.8 or higher only.
-* Rename all APIs to use "session" instead of "kernel" ([#216](https://github.com/lablup/backend.ai-manager/issues/216))
-* API v5 support
-  - All session-related GraphQL schema is rewritten to reflect multi-container sessions and kernel -> session parameter and API renamings. ([#263](https://github.com/lablup/backend.ai-manager/issues/263))
-
-### Features
-* Un-numbered features
-  - Now all CLI commands are accessible via ``backend.ai mgr``. [(lablup/backend.ai#101)](https://github.com/lablup/backend.ai/issues/101)
-  - Add more convenient manager CLI commands: `etcd quote`, `etcd unquote`, `etcd move-subtree`
-  - Add `--short` and `--installed` options to the `etcd list-images` command.
-  - Add a minimum-occupied slot first scheduler.
-* Now our manager-to-agent RPC uses [Callosum](https://github.com/lablup/callosum) instead of aiozmq, supporting Python 3.8 natively. ([#209](https://github.com/lablup/backend.ai-manager/issues/209))
-* Add vfolder large-file upload APIs using the tus.io protocol. ([#210](https://github.com/lablup/backend.ai-manager/issues/210))
-* User-customizable per-scaling-group session queue scheduler plugins and three intrinsic plugins: FIFO, LIFO, and the DRF (dominant resource fairness) scheduler. ([#212](https://github.com/lablup/backend.ai-manager/issues/212))
-* User-manageable session templates written in YAML to reuse session creation parameters. ([#213](https://github.com/lablup/backend.ai-manager/issues/213))
-* ResourceSlots are now more permissive so that agents with different sets of accelerator plugins can now coexist in a single cluster. ([#214](https://github.com/lablup/backend.ai-manager/issues/214))
-* Add pre-open service ports to allow user-written applications listening on a container port and make such ports accessible via the stream proxy API. ([#221](https://github.com/lablup/backend.ai-manager/issues/221))
-* Auto-fill the minimum resource slots for intrinsic slots only to allow execution of sessions with different set of accelerators simultaneously. ([#234](https://github.com/lablup/backend.ai-manager/issues/234))
-* Error logging API and an intrinsic plugin to store all unhandled exceptions in agents and managers.
-  (Agent exceptions are passed via the event bus) ([#235](https://github.com/lablup/backend.ai-manager/issues/235))
-* Reduce possibility of aioredlock locking errors. ([#236](https://github.com/lablup/backend.ai-manager/issues/236))
-* Favor CPU agents for session creation requests without accelerators ([#241](https://github.com/lablup/backend.ai-manager/issues/241))
-* Add user-config APIs to update/get keypair bootstrap script ([#253](https://github.com/lablup/backend.ai-manager/issues/253))
-* Accept session UUIDs and their prefixes in addition to client-specified IDs (names) ([#258](https://github.com/lablup/backend.ai-manager/issues/258))
-* API v5 support ([#263](https://github.com/lablup/backend.ai-manager/issues/263))
-  - Add full implementation of the background-task tracking API
-  - Apply background-task tracking API to the image rescanning API
-  - Add paginated UserList and KeyPairList GraphQL queries
-* Extend vfolder to have usage mode and innate permission ([#265](https://github.com/lablup/backend.ai-manager/issues/265))
-* Add shared memory column to resource preset ([#267](https://github.com/lablup/backend.ai-manager/issues/267))
-
-### Fixes
-* Fix regression of the docker registry list fetch API (`/config/docker-registries`) ([#259](https://github.com/lablup/backend.ai-manager/issues/259))
-* Fix errors on fetching server logs and vfolder list ([#264](https://github.com/lablup/backend.ai-manager/issues/264))
-* Remove unwanted empty key while setting etcd config with nested dict. ([#275](https://github.com/lablup/backend.ai-manager/issues/275))
-
-### Miscellaneous
-* Internally refactored the main function for easier writing of future unit tests by composing different resource cleanup contexts in a modular way, using aiohttp's cleanup contexts.
-* Now all changelogs are managed using [towncrier](https://github.com/twisted/towncrier) and migrated to Markdown syntax. ([#262](https://github.com/lablup/backend.ai-manager/issues/262))
-* Refactor `batch_result()` and `batch_multiresult()` used for optimizing SQL-based GraphQL queries ([#263](https://github.com/lablup/backend.ai-manager/issues/263))
-* Return NaN when group-level resource status is not allowed to the client. ([#268](https://github.com/lablup/backend.ai-manager/issues/268))
-* Update flake8 to a prerelease version supporting Python 3.8 syntaxes ([#277](https://github.com/lablup/backend.ai-manager/issues/277))
diff --git a/changes/409.fix b/changes/409.fix
deleted file mode 100644
index f83032dbd..000000000
--- a/changes/409.fix
+++ /dev/null
@@ -1 +0,0 @@
-Fix a missing reference fix for renaming of `gateway` to `manager.api`
diff --git a/changes/410.misc b/changes/410.misc
deleted file mode 100644
index 0ec717292..000000000
--- a/changes/410.misc
+++ /dev/null
@@ -1 +0,0 @@
-Fix the examples for the storage proxy URL configurations in the `manager.config` module
diff --git a/changes/411.fix b/changes/411.fix
deleted file mode 100644
index 15ad49f10..000000000
--- a/changes/411.fix
+++ /dev/null
@@ -1 +0,0 @@
-Refactor the manager CLI initialization steps and promote `generate-keypair` as a regular `mgr` subcommand
diff --git a/changes/412.fix b/changes/412.fix
deleted file mode 100644
index 2f1a469b6..000000000
--- a/changes/412.fix
+++ /dev/null
@@ -1 +0,0 @@
-Fix an internal API mismatch for our SQLAlchemy custom enum types
diff --git a/changes/413.fix b/changes/413.fix
deleted file mode 100644
index a23a11f8b..000000000
--- a/changes/413.fix
+++ /dev/null
@@ -1 +0,0 @@
-Fix a regression in session cancellation and kernel status updates after SQLAlchemy v1.4 upgrade
diff --git a/changes/414.misc b/changes/414.misc
deleted file mode 100644
index 45ccb2ef0..000000000
--- a/changes/414.misc
+++ /dev/null
@@ -1 +0,0 @@
-Update sample configurations for etcd
diff --git a/src/ai/backend/manager/__init__.py b/src/ai/backend/manager/__init__.py
index 1f87871fd..7467c3b57 100644
--- a/src/ai/backend/manager/__init__.py
+++ b/src/ai/backend/manager/__init__.py
@@ -1 +1 @@
-__version__ = '21.03.0'
+__version__ = '21.03.1'