- Bug in installing SLURM by means of spack-based installation
- Setting of parallel branches at runtime #68
- Bug in reporting the output of light tasks of running workflows
- Bug causing a possible buffer overflow
- Parsing input data as a single command: only JSON Requests are admitted
- Script to start Ophidia I/O Servers by means of OPH_CLUSTER
- Handling of flag 'output' in setting its value to an input argument from a dependence
- Port in template of Ophidia IO Server configuration
- Argument 'fileref' to OPH_WAIT #64
- New argument ENABLE_CLUSTER_INCREASE
- Argument AUTO_RETRY is set to 0 by default
- Bug in parsing for comments in JSON Requests
- Possibility to create one branch for each file of a folder with parallel OPH_FOR #63
- Submission control to limit the number of running tasks or busy cores #62
- Simple scripts to generate digital certificates
- Workflow variable indicating the number of loops of a 'OPH_FOR' statement
- Save last commands by name in case of long strings
- Enable the possibility to use '=' in parameter values
- Support for CentOS9
- Support to show information about datacubes and containers in the web interface
- Program soapcpp2 is necessary only to make oph.wsdl, but not to build the package
- Script and configuration to start I/O servers using OPH_CLUSTER
- Bug in reading the global option 'checkpoint' #60
- Option 'direct_output' to control the output of single-task workflows #61
- Bug in handling the ordered list of output cubes
- Bug in OPH_CLUSTER to return the JSON result in case of wrong input argument
- Bugs in configuration script for code building
- Bugs in workflow engine
- Procedure to evaluate values of parameters passed from a task to another one
- Execution of OPH_RESUME in case of request for a specific task
- Allocation of JSON Requests to heap memory for comment deletion
- Bug in evaluation of user quota in OPH_CLUSTER
- Bug in debug printing function
- Support for cube argument for importncs #58
- Features to load/store variables from/to JSON Requests
- Support for checkpointing workflow execution
- Dynamic back-off in case of task re-submission
- Improve reliability of OPH_WAIT in case of cancellation
- New 'measure' and subset parameters to OPH_WAIT operator
- 'author' and 'abstract' are not mandatory parameters of a workflow anymore
- Bug in handling the arguments 'enable' and 'disable' of OPH_SERVICE
- Bug in transfer metadata into accounting table
- Bug in handling cube pids for final operation
- Option to enable and disable JSON response saving #56
- New argument 'WORKING_DIR' in main configuration file
- File check of OPH_WAIT from 'stat' to OPH_FS #57
- Activate service without authorization procedure in case authorization directory is not found
- Bug in formatting submission strings
- Order of the arguments 'workflowid' and 'project' of oph_submit.sh
- Bug in parsing the arguments 'values' and 'counter' of OPH_FOR
- Bug in handling country name while using the tool oph_manage_user
- Bug in parsing multivalue arguments
- Typo in OPH_CLUSTER operator XML file
- Code definition related to OpenID Connect
- Bug in processing configuration file
- Bug related to requests where sessionid is set to NULL
- Server core functions to support longer input strings for commands and workflows
- OPH_MANAGE_SESSION when action 'new' is called
- Use unique job identifiers also for internal requests associated with massive operations #47
- Check for argument 'nthreads'
- New argument 'ENABLE_CANCEL_ALL' in main configuration file #55
- Field 'project' to header of JSON Request #54
- Basic JSON output for workflows and tasks aborted by OPH_CANCEL
- Argument 'exec_mode' to OPH_WAIT
- Monitor of the number of active threads in server #53
- Options to generate random user passwords and execute post-scripts to set env variables #52
- Update of task status in OphidiaDB at the end of execution
- New argument 'OPENID_ALLOW_OFFLINE_ACCESS' in main configuration file
- New argument 'ENABLE_AUTHORIZATION' in main configuration file and option 'a' to disable authorization procedure
- Support to retrieve username from email address
- Check of email provider as organisation name for OpenId Connect
- Systemd unit file for the server
- Sync execution mode when oph_cluster is used for deploying
- Creation date of host partitions to output of OPH_CLUSTER #50
- Feature to enable and disable users selectively #49
- New options in configuration file to set OpenID username as Ophidia user #48
- Status log of the workflows started and completed in last period
- Polling procedure for aborted or starved tasks
- Support for OpenId Connect #51
- OPH_SET and OPH_INPUT to allow 0 as a value for the 'id' argument
- Log system in order to provide additional information regarding tasks/workflows
- Host partion option in worklow specification to support 'auto' keyword
- Check for arguments 'ncores' and 'nhost'
- Useless variables from oph_plugin
- Status OPH_ODB_STATUS_CLOSED from list of possible states
- Support for massive operations on parameter 'cubes' #46
- Accounting features to log a line for each massive task #45
- Default value of parameter 'action' of OPH_MANAGE_SESSION
- Issue #41
- Bug #33
- Bug in reading rmanager configuration file with empty lines
- Bug in update OPH_USER_LAST_CDD by means of oph_manage_user
- Name of pre-defined runtime variable NHOST
- New configuration option OPH_OS_USERNAME to define the OS user to be used for command submission #40
- Arguments to OPH_CLUSTER to get info about the host partitions #38
- A new resource manager command to stop host partition #37
- Arguments to specify subsets in OPH_SET #35
- Support for OPH_DELETECONTAINER as workflow exit action #34
- Soap interface files with gSOAP version 2.8.76 #42
- Default value of argument nhost of OPH_CLUSTER
- Host partition management according to OphidiaDB refactoring #36
- Output of OPH_SERVICE
- Raise an error in case any task parameter contains special characters
- References to hidden containers #39
- Parameter 'nhost' to JSON Request schema
- Feature for dynamic cluster deployment #31
- New argument 'ENABLE_CLUSTER_DEPLOYMENT' in main configuration file #31
- Clear intemediate results from job table in case of unexpected failures
- Support to extract arrays from JSON Responses and to process them as run-time variables #29
- Extend the set of pre-defined run-time variables #30
- Bug in handling negative filters
- Bug in handling Ophidia DB access failures
- Bug in postprocessing oph_system
- Unit tests
- Bug in workflow comments parsing
- Improvement of support for large variables #32
- OPH_CLUSTER xml file #31
- slurm.conf configuration file arguments #31
- Removed slurm_multiuser.conf configuration file #31
- oph_manage_user to set initial OPH_CDD argument when user is first created
- Keyword 'end' to refer to JSON fields using OPH_SET #27
- Support to create, start and terminate user-reserved host partitions and associated hosts - DBMS instances #26
- Extra fields in workflow response #25
- Support for negation in filters for massive operations #24
- Possibility to evaluate expressions in OPH_SET #23
- Support for comments in workflow request #22
- Some debug messages in workflow validation function
- Algorithm used to check workflow validity.
- Checking for libssh when it is not enabled in configuration
- Job tracking for accounting purposes #28
- Monitoring log to include number of incoming/closed tasks during a period
- Information shown in accounting log
- Code coverage macros to allow the use of LCOV 1.13
- Unit tests
- Bug #13 related to large variables causing server crash when used in OPH_SET #21
- Bug #20
- Update resource manager to set specific MPI type in case only one core is used #16
- Bug #14
- Bug #12
- Improve metadata management to prevent deadlocks #11
- Bug #10
- Bug #9
- Locking for libcurl
- Add accounting log #19
- Feature AAAaaS #17
- URL to workflow header #15
- Workflow execution time as extra field
- Fields 'category' and 'multivalue' to XML descriptions
- Default values for configuration parameters of OpenId Connect
- OPH_CDD as parameter of user.dat and output of OPH_GET_CONFIG
- Timer to check for revoked tokens
- Update fill partition query to use datacubecount field #18
- Look of web pages of session space
- Use subject identifier as userid for OpenId Connect
- Default values for configuration parameters
- Unit tests
- Bug #8
- Bug #7
- Limit the allowed characters for 'username'
- OPH_FOR to avoid to re-check the workflows in case the option "parallel" is not used
- OPH_WAIT to check files
- Fix bug in handling on_error flag in loops
- Support for OpenId Conncect (INDIGO-DataCloud Project)
- Add BASE_SRC_PATH in client configuration parameters
- Fill extra metadata in responses of known operators (INDIGO-DataCloud Project)
- Multi-user support
- Add fine progress ratio evaluation based on number of fragments
- Add status logging and monitoring feature
- Disable SSH for task submission by default
- Add sessionid and markerid to success notifications
- Import massive operations based on the operator OPH_FS (INDIGO-DataCloud Project)
- JSON Response schema (version 1.1): add field 'extra' (INDIGO-DataCloud Project)
- OPH_RESUME to get workflow progress
- OPH_IF to forward input cube (INDIGO-DataCloud Project)
- Allow embedded variables
- Allow to use @ even if no variable is not set
- Save JSON Responses on behalf of the framework
- Trace operator log only in debug mode
- Localtime function calls to be reentrant
- Unit tests
- Bug that raises an error message at the end of leaf tasks
- Bug when a waiting task is terminated by OPH_CANCEL
- Bug in job monitoring procedure
- Bugs in building the list of cubes to be processed by 'exit action'
- Use of BASE_SRC_PATH as prefix of files involved in massive and waiting operations (INDIGO-DataCloud Project)
- Parameter 'status_filter' to OPH_RESUME
- Status OPH_RUNNING_ERROR in workflow in case of error while it is still running
- Disabled usage of libSSH by default
- Exit status saved in OphidiaDB when workflow is aborted
- System command in case libSSH is not used
- Number of cores to execute each light task of final task to 1
- SSH connection method
- OPH_SERVICE to list running tasks
- Configuration parameters names in ophidiadb.conf
- Bug in testing OPH_WAIT (INDIGO-DataCloud Project)
- Bug #6
- Option 'host_partition' in workflow schema
- Improve code coverage of unit tests (INDIGO-DataCloud Project)
- Memory leaks in testing program
- Inclusion of curl.h in code coverage (INDIGO-DataCloud Project)
- Warning messages set in JSON response in case of successful tasks
- Bug in handling wrong expression with OPH_IF (INDIGO-DataCloud Project)
- Bug in setting exit status of OPH_ENDFOR (INDIGO-DataCloud Project)
- Compact output format in resuming status of running workflow
- Improved procedure to set final operation arguments
- Bug in OphidiaDB update when a massive operation is retried
- Bug in handling OPH_IF when Matheval is not enabled (INDIGO-DataCloud Project)
- Bug in job monitoring procedure
- Bug in auto-retry feature
- New unit tests for OPH_WAIT, OPH_INPUT and OPH_SET (INDIGO-DataCloud Project)
- Htacces to define url redirect for Ophidia web section
- Argument to define type of operation in OPH_CANCEL
- BASE_BACKOFF configuration parameter
- BASE_SRC_PATH configuration parameter (INDIGO-DataCloud Project)
- New operator OPH_INPUT (INDIGO-DataCloud Project)
- New operator OPH_WAIT (INDIGO-DataCloud Project)
- Query of OPH_RESUME to return OPH_STATUS_WAITING in case there is at least a waiting task (INDIGO-DataCloud Project)
- Workflow status list; added OPH_STATUS_WAITING and OPH_STATUS_UNSELECTED (INDIGO-DataCloud Project)
- Policy to set auto-retry interval (backoff-based policy)
- OPH_FOR and OPH_SET input parameter "name" to "key"
- Default value for SERVER_FARM_SIZE to 128
- Handler of SSH library used for task submission, locking, sessions, etc.
- Bug in setting basic notification message for massive operations (INDIGO-DataCloud Project)
- Skip 'lib' in referring to libmatheval
- Job status when OPH_START_ERROR occurs
- Operation to be executed when OPH_START_ERROR occurs in case of the final task
- Counter update of OPH_FOR when default values are used (INDIGO-DataCloud Project)
- Bug in building function 'oph_if_impl' when MathEval library is disabled (INDIGO-DataCloud Project)
- Bug in handling filter 'path' in massive operations (INDIGO-DataCloud Project)
- Bug in massive operation handler (INDIGO-DataCloud Project)
- Several warnings when building
- Monitor of the submitted jobs
- Support for auto-retry in case of resource manager errors
- Check for return values of SSH-submitted commands
- Compact output format for JSON Responses associated with workflows and massive operations (INDIGO-DataCloud Project)
- Task type OPH_STATUS_UNSELECTED for unselected branches of selection statement (INDIGO-DataCloud Project)
- Code coverage check (INDIGO-DataCloud Project)
- Support for data type 'string' in multigrids of JSON Responses
- Unit tests (INDIGO-DataCloud Project)
- Support for selection statement (INDIGO-DataCloud Project)
- Status of skipped tasks to OPH_STATUS_SKIPPED in case of errors (INDIGO-DataCloud Project)
- Code indentation style (INDIGO-DataCloud Project)
- Library 'known operators' to improve modularity (INDIGO-DataCloud Project)
- Version number in files
- Bug in OPH_FOR parsing
- Warning 'unset variable' in oph_ssh_submit.c
- Bug in function for massive operation checking
- Bug in file matching during directory scanning
- Bug in handling subset strings for argument 'counter' of OPH_FOR
- Bug in handling on_exit:oph_delete in case of massive operations
- Bug #4
- Bug #2
- Bug #1
- Support for pre-defined workflow variables
- Support for hostname in IP_TARGET_HOST configuration parameter
- Support for building on CentOS7 and Ubuntu
- Handling of DT_UNKNOWN in parsing directories for import massive operations
- Support for SHA-coded passwords
- Improved oph_for and oph_set to handle vectors as symbolic references
- Initial public release