Commons io

DataSpaceServiceStarter: username subfolder should only be created wh…

Co-authored-by: test <test@test.com>

DataSpaceServiceStarter: remove obsolete localpath check

improve LoadPackage ssl handling

TrustSelfSignedStrategy fails in some cases where TrustAllStrategy should work in all cases

SchedulerNodeClient: when connection fails, use the scheduler rest pu…

When connection to the scheduler fails (for example, when the scheduler is behind a reverse proxy), and a scheduler rest public url is configured, the scheduler client will attempt to connect using the public url.

* Add jobId filter to GET jobrevision

* Add comment in endpoint definition + spotlessApply

* Make param final getJobs method

Add Start At filter for GET revisionAndJobsInfo endpoint

Remove config of Start At as an extra column. It is now a field in JobInfo, read by the scheduler portal directly (#4288)

Add description to JobInfo

Add startAt to JobData and Jobinfo

startAt is stored in the database and searchable

ensure that changeStartAt request updates the value stored in JobData and the generic info

add a changeStartAt request that accepts multiple jobs

Fix updateStartAt

When updating the startAt value of a pending job, the cache used by ExtendedSchedulerPolicy need to be updated.

Accordingly, the StartAtUpdater must be able to retrieve the policy instance and ask it to update the startAt value of a given job or task.

Improve performance of ExtendedSchedulerPolicy

 - use an LRU cache for getStartAtValue (this prevents costly operations over the same job and task)
 - changed the key of delayedJobsWakeUpCache to be the startAt time instead of job id
 - added timer loggers to monitor performances when the logger is in debug

* Refactor POST and DELETE credentials to take key as form param. Remove @Encode annotation and manual decode() for automatic decoding of key.

* Accept credential key as QueryParam for DELETE

Fix catalog portal id in portals access

improve the job owner check

instead of using JaaS, simply compare the current user's name and tenant with the job owner's name and tenant

* Prevent duplicates when creating labels. Also ensure front-end backward compatiblity by returning unique results when fetching labels from DB that have same label name

* Remove duplicates when creating labels

* Throw Exception if duplicate labels are found

* Refactoring: Make check case sensitive + do not throw exception, instead, remove duplicates

* Apply comment: Use LinkedHashSet to preserve order

Fix typo in rest api url

fix issues in NodeUsageSchedulingPolicy

 * when a job is finished, if it's a top level job, wait until all children jobs are terminated before removing all knowledge of the job.
 * remove finished tasks before removing finished jobs.

add NodeUsageSchedulingPolicy