## Summary

This release introduced an additional gRPC based frontend for storage backends. This is still a work in progress and is not ready for use.

The command `cta-verify-file` now requires the options `eos.instance`, `eos.request.user` and `eos.request.group` to be configured in `/etc/cta/cta-cli.conf`.

### Upgrade Instructions
### Features
- cta/CTA#1222 - Add minimal gRPC based frOntend for integration with dCache
- cta/CTA#1241 - make cta-verify-file get instance, request user and group otions from ctacli config file
### Bug fixes
- cta/CTA#1225 - Fix bug causing tapeserver to sometimes pop the entire archive queue at the end of the mount
### Building and Packaging
### Catalogue Schema

### Features
- cta/CTA#1205 - Fail pipeline if cppcheck detects errors
- cta/CTA#1206 - Change NULL for nullptr
- cta/CTA#1217 - Schema verification should just issue an warning if there are extra indexes in the db that are not in the catalogue
- cta/CTA#1220 - Improve queued retrieve logging message
- cta/CTA#1152 - Reduce eos free space query load
- cta/CTA#1231 - Add --loadtimeout option to cta-tape-label and increase default value
- cta/CTA#977 - Add --drive option to cta-tape-label

### Bug fixes
- cta/CTA#1120 - Fix negative disk space reservation content
- cta/CTA#1235 - cta-verify-file should return 1 on error

4.6to10.0 migration script skips creating index ARCHIVE_FILE_SCI_IDX if it already exists

## Summary
### Upgrade Instructions
### Features
- cta/CTA#1161 - Tape server refactoring, "Decide where m_reportPacker.setTapeDone() should be called"
- cta/CTA#1195 - `cta-catalogue-schema-drop` should drop the CTA_CATALOGUE table last
### Bug fixes
- cta/CTA#950 - Eliminate race condition preventing the drive to go down on failure
- cta/CTA#1160 - Fix DrainingToDisk stale status in case if there is DiskWriteThreadPool thread left
- cta/operations#708 - Fix "Should run cleaner but VID is missing. Putting the drive down"

### Building and Packaging
### Catalogue Schema

## Summary
Tape server refactoring, improvements and bug fixes

### Upgrade Instructions
### Features
- cta/CTA#1161 - Tape server refactoring, "Decide where m_reportPacker.setTapeDone() should be called"
### Bug fixes
- cta/CTA#950 - Eliminate race condition preventing the drive to go down on failure
- cta/CTA#1160 - Fix DrainingToDisk stale status in case if there is DiskWriteThreadPool thread left
- cta/operations#708 - Fix "Should run cleaner but VID is missing. Putting the drive down"

### Building and Packaging
### Catalogue Schema

## Summary
### Upgrade Instructions
### Features
- cta/CTA#1179 - Use std::optional instead of cta::optional
- cta/CTA#1190 - Use std::make_unique instead of cta::make_unique
- cta/CTA#1198 - Use hardcoded mount policy for verification requests
- Add verification flag to queued retrieve request log message
- propagate labelFormat from TAPE catalog to VolumeInfo
- cta/CTA#1200 - Remove range class

### Bug fixes
### Building and Packaging
### Catalogue Schema

This release upgrades the catalogue schema to version 10.0, which removes remove unused tables, introduce new indexes and foreign key constraints.

Patch to fix stale drives and wrong data movement timeout

Patch to fix stale drives

## Summary

### Upgrade Instructions

This CTA release requires a database schema upgrade to CTA catalogue schema v4.6.
Please consult the [database upgrade documentation](https://tapeoperations.docs.cern.ch/ctaops/upgrade_production_database).

### Features
- cta/CTA#1137 - Stop deletion of failed retrieve/archive requests
- cta/CTA#1150 - Add option to pass desired catalogue version into cta-catalogue-schema-create
- cta/CTA#1147 - Add Disk Instance Column to VO table
- cta/CTA#1119 - Remove support of manual mode for loading tapes
- cta/CTA#1123 - Add mount id to disk space reservations, prevent tape servers from releasing disk space from a previous mount

### Bug fixes
- cta/CTA#1138 - sortAndGetTapesForMountInfo only queries tapes in the current logical library
- cta/CTA#1117 - Update masterDataInBytes when writing files to tape
- cta/CTA#1125 - cta-admin dr ls should show '-' instead of "NO_MOUNT" for Mount Type

### Continuous Integration
- cta/CTA#1131 - Compile CTA using devtoolset-8 in CI
- cta/CTA#1126 - Create tests for TapeDrivesCatalogueState

- Workaround for negative reservation crashing the tapeserver and leaving the tape stuck

- cta/CTA#1101 - Fix disk space reservation logic adding all existing disk space reservations for all disk systems
- cta/CTA#1023 - Retrieve puts the queue to sleep if the eos disk instance is not reachable

## Features
- cta/CTA#1091 - New Drive State table in CTA Catalogue
- cta/CTA#1054 - Fix filing of disk buffer when recalling from tapeservers with RAO
- cta/CTA#1076 - Retrieve fails if disk system configuration is removed
- cta/CTA#1087 - Add new tapeserver config option UseEncryption
- cta/CTA#1096 - Better handling of bad checksums in archive/retrieve sessions

## Bug fixes
- cta/CTA#1092 - Fix overflow error with drive state latestBandwith causing cta frontend crash
- cta/CTA#501 - Fix disappearing reason when TapeDrive is reading or writing

## Bug fixes
- cta/CTA#1092 Fix overflow with drive bandwith causing frontend to crash

- cta/CTA#1053 Remove the objectstore presence in another classes
- cta/CTA#501 cta-taped should set the state to DOWN when machine rebooting
- cta/CTA#1056 Fix bugs in cta 4.3-2"
- cta/CTA#1058 Remove helgrind_annotator from production

- cta-admin: All short options with more than one character now require two dashes
- cta/CTA#1013 reportType uninitialized
- cta/CTA#1044 Fix segmentation fault due to uninitialized optional value and remove diskSpaceReservations from `cta-admin dr ls`

- cta/CTA#983
  - Add cta-versionlock helper script to cta-release package
  - Update cta repo file to use the new public repo
- cta/CTA#1036 Better error reporting in cta-admin tools
- cta/CTA#1039 Improve logging of cta admin commands in cta frontend
- cta/CTA#1041 Fix host values in cta-admin commands
- cta/CTA#1026 `cta-admin sq` now shows the mount policies of potential mounts
- cta/CTA#1029 Fix segmentatin fault in frontend when list repacks of a tape that has been deleted in the catalogue
- cta/CTA#1032 cta-admin dr ls crashes the frontend if executed during an archive/retrieve
- cta/CTA#996  Removes PARALLEL from migration scripts
- cta/CTA#1035 log configuration options on frontend startup
- cta/CTA#1042 Do not iterate over retrieve queues when holding global scheduler lock
  - Repacks on a disabled tape must now use a mount policy whose name starts with repack
  - There is no longer an empty mount when a disabled/broken tape queue is full of deleted requests
- cta/CTA#1027 Mitigate popNextBatch bad behaviour in archive queues
- Migration tools: fixes filemode of files imported from CASTOR to EOS

## Summary

### Features
- cta/CTA#501 cta-taped should set the state to DOWN when machine rebooting
- cta/CTA#976 Add logical part of Drive Status using Catalogue
- cta/CTA#983
  - Add cta-versionlock helper script to cta-release package
  - Update cta repo file to use the new public repo
- cta/CTA#988 Instrument backpressure for retrieve in dr ls
- cta/CTA#1036 Better error reporting in cta-admin tools
- cta/CTA#1039 Improve logging of cta admin commands in cta frontend
- cta/CTA#1041 Fix host values in cta-admin commands

### Bug fixes
- cta/CTA#1031 Fix Warning in updateDriveStatus
- cta/CTA#1032 cta-admin dr ls crashes the frontend if executed during an archive/retrieve
- cta/CTA#996  Removes PARALLEL from migration scripts
- cta/CTA#991  Drive is not put down if the device file is removed while cta-taped is running and before a data transfer