Tags

- [frontend] Add options to "tapepool ls" to filter tapepools on their name, vo and encryption
- cta/CTA#898 cta-send-event now gets the requester id and eos instance as command line arguments
- cta/CTA#1005 "tape ls" now can filter tapes on wether they were imported from CASTOR
- cta/CTA#1006 "repack ls" now shows the tapepool of the tape being repacked
- [frontend] Adds missing break after "schedulinginfo ls" command
- cta/CTA#999 Adds a default mount rule
- cta/CTA#1003 The expansion of a repack request now fails if the archive route for archiving the repacked files is missing

Bug fix for infinite error reporting retry

- Upgraded ceph to version 14.2.20
- Adds cta-verify-file to cta-cli RPM

- Upgraded EOS to 4.8.45
- Upgraded eos-xrootd to 4.12.8
- cta/CTA#966 Unable to distinguish empty and wrong tape pool
- cta/CTA#926 Improve MigrationReportPacker::ReportSkipped::execute() exception message
- cta/CTA#584 Validate checksum when recalling from tape

* Fri Feb 05 2021 julien.leduc (at) cern.ch - 4.0-1
- cta/CTA#964 Adds failure log messages to processCLOSEW in CTA Frontend
- When the operator submits a tape to repack, a check is done about the tape state before queueing the repack request to ensure it can be repacked
- Oracle catalogue migration scripts 3.1to3.2.sql: replaced DELETE FROM table_name by TRUNCATE TABLE table_name

* Wed Feb 03 2021 julien.leduc (at) cern.ch - 3.2-1
- Upgraded EOS to 4.8.37-1
- cta/CTA#922 The superseded concept has been removed and replaced by a new recycle bin
- cta/CTA#943 A new tape lifecycle logic has been implemented
- cta/CTA#948 The max drives allowed for reading and writing are now set per virtual organization and not per mount policy anymore
- cta/CTA#952 Reclaiming a tape resets the IS_FROM_CASTOR flag to 0
- cta/CTA#951 The query used by RdbmsCatalogueTapeContentsItor has been put back to the state it was in v3.1-13
- cta/CTA#883 Tape verification tool

- Upgraded EOS to 4.8.35-1
- cta/CTA#954 The drive is put down if the CleanerSession fails to eject the tape from it after a shutdown
- cta/CTA#945 The timeout for tape load action is now configurable
- cta/CTA#957 The state of the successful Archive for repack jobs is now changed before being queued for reporting
- cta/CTA#958 The RepackRequest garbage collection now changes the owner of the garbage collected RepackRequest.

* Thu Jan 07 2021 julien.leduc (at) cern.ch - 3.1-13
- Upgraded EOS to 4.8.34-1
- Upgraded xrootd to 4.12.6-1
- cta/CTA#941 Slow `cta-admin sq` even when there is very little activity
- cta/CTA#951 Improve the performance of RdbmsCatalogueTapeContentsItor
- cta/CTA#939 cta-objectstore-dereference-removed-queue removes all kind of manually deleted queues from the RootEntry

- Upgraded EOS to 4.8.30-1

- cta/CTA#932 Add environment file for cta-frontend service: frontend configured to use 10 XRootD polling threads by default
- cta/CTA#292 Allow non interactive usages of cta-admin with sss authentication
- Upgraded EOS to 4.8.29-1
- cta/operations#155 Fix for conversion issues
- cta/operations#154 Improve sys.retrieve.req_id to allow to cancel retrieves on a running instance: adding epoch timestamp in ids
- EOS-4505 Separate archive and retrieve ACLs in EOS: only needs p ACL for prepare
- cta/operations#150 high priority Archive job not scheduled when Repack is running: fixed

* Thu Nov 12 2020 julien.leduc (at) cern.ch - 3.1-10
- cta/CTA#837 Repack now fails if the repack buffer VID directory cannot be created during expansion
- cta/CTA#920 Archive and Retrieve error report URL correction on the cta-send-event cmdline tool
- cta/CTA#923 Corrected the cta-admin showqueues command to display all the retrieve queues of tapes that are on the same tapepool
- Unuseful WARNING logs are now DEBUG logs

* Wed Nov 04 2020 julien.leduc (at) cern.ch - 3.1-9
- Upstream EOS 4.8.26-1
- cta/CTA#907 For backpressure, the EOS free space can be fetched by calling an external script
- cta/CTA#917 Corrected the bug in the cta-admin showqueues command in the case ArchiveForUser and ArchiveForRepack exist for the same tapepool
- cta/CTA#919 Archive queue oldestjobcreationtime is now updated at each pop from the ArchiveQueue

- CTA software Recommended Access Order (RAO) implemented for LTO drives
- cta-admin repack ls tabular output improvements
- Upstream EOS 4.8.24-1
- Upstream xrootd to 4.12.5-1
- Repack management execution can be disabled via the cta-taped configuration file
- cta/CTA#907 Maintenance process can be disabled via the cta-taped configuration file
- Catalogue refactoring
- utils::trimString() now returns an empty string if the string passed in parameter contains only white-space characters
- cta/CTA#895 [catalogue] RdbmsCatalogue::deleteLogicalLibrary does not delete empty logical library
- Repack request and sub-requests are now unowned from their Agent when completed

- cta/CTA#893 Corrected slowliness of RdbmsCatalogue::getArchiveFileToRetrieveByArchiveFileId()

- cta/CTA#881 cta-fst-gcd logs can be now sent to stdout by command line option for container based deployments
- cta/CTA#885 cta-admin should be able to query by sys.archive.file_id
- Upstream eos 4.8.15-1
- Upstream xrootd 4.12.4-1
- cta/CTA#890 CTA RPMs should only use the xrootd-client-libs package
- buildtree installation scripts are made compatible with Centos 7
- cta/CTA#892 Modified the log level of the triggering of Archive and Retrieve mounts
- cta/CTA#889 It is not possible to retrieve a file that is not active anymore
- cta/CTA#877 ObjectStore.RetrieveQueueAlgorithms unit tests fails or succeeds base on version of cmake
- cta/CTA#888 Garbage collector race condition
- cta/CTA#891 Corrected Repack Archive subrequest creation time

- cta/CTA#863 Prevent SQLite database files from being used as the CTA catalogue database backend
- cta/CTA#861 cta-admin comment column is flush left
- cta/CTA#862 Unable to delete tabtest tape pool because it is in an archive route
- cta/CTA#860 Correct contents of cta-lib-catalogue RPM and correct dependencies on it
- Reinstates "cta-admin failedrequest --summary" option
- cta/CTA#865 Empty the RetrieveQueue in the case of cancellation of a retrieve request when the drive is down

Improvements to cta-admin and data performance plus removal of deprecated cta-objectstore-unfollow-agent