Cleaning up run scripts for containers: going along Dockerfile and in /opt/run/bin: /shared still available for local scripts developments.

DOCKER_LOGIN_USERNAME
DOCKER_LOGIN_PASSWORD
set in /etc/gitlab/gitlabregistry.txt managed by Puppet

This is needed because gitlabregistry does not implement search.
therefore, one needs to list images through gitlab curl API using JSON
Web Token.
This is needed for tests outside of CI, were one must deduce ctageneric
docker image tag from commit ID, the format is <pipeline ID>git<commit
ID> and while we have access to commit ID, we cannot guess pipeline ID.
The idea is then to run the tests with the highest pipeline ID and the
commit ID of HEAD.

simplify default run.sh because we simply do not use it. Does nothing but waits for pod termination...

The log line creation function Logger::writeLogMsg does not add the trailing newine.
Instead, the overloads of reducedSyslog() do, il needed and after possible line truncation.

Added remarks that poped up after the meeting (SSI and threading model, security, CASTOR migration and disaster recovery).

prio set at 4 for jenkins EOS repo to allow tests against the latest EOS

Improved the wording related to synchronous vs asynchronous as both could be implemented with EOS's WFE.

Added a high level description of the EOS-CTA interface derived from the consideration of user access paterns.

This doc is now included in cta.tex. The standalone ObjectStore.pdf should have been deleted at that time.

This new package contains the old makeMinimalVFS (renamed to cta-objectstore-initialize)
plus cta-objectstore-list and cta-objectstore-dump-object.

This was a slight variation of the previous case (again a complain on mutex_destroy).

https://st-ci.cern.ch/job/CTA-autotest/1034/consoleFull

Fix to #47: OStoreDB::getArchiveJobs() now returns an empty list for tape pools unknown to object store.

The first 6 issues are about places where error conditions of the pthread mutexes are validated (And a suppression is required).
The 7th one is about a race between the unlocking of the mutex itself and it deletion. We seupresed this error also.

This commit works towards fixing the following issue:

https://gitlab.cern.ch/cta/CTA/issues/46
issues when creating 100+1000 files on eos-cta via Kubernetes setup