Test thumbs

[pwalczysko]

What this PR does

Fixing the failure of 2 tests, one at https://merge-ci.openmicroscopy.org/jenkins/job/OMERO-test-integration/lastCompletedBuild/testReport/OmeroPy.test.integration.test_thumbs/TestThumbs/testThumbnailVersion_one_/ and another of https://merge-ci.openmicroscopy.org/jenkins/job/OMERO-test-integration/lastCompletedBuild/testReport/OmeroPy.test.integration.test_thumbs/TestThumbs/testThumbnailVersion_set_/

• test/integration/test_thumbs2.py::TestThumbs::testThumbnailVersion[one]
• test/integration/test_thumbs2.py::TestThumbs::testThumbnailVersion[set]

Again, the migration to NFS necessitates wait for pyramid gen and check via checking of the filesystem itself.
Reusing the method introduced in #6464

Testing this PR

Check Jenkins is green for that test.

Related reading

Similar problematics as #6464

[pwalczysko]

The added waits for the pyramid building are working - but unfortunately, the total situation of testintegration on Jenkins is not improved by that.

1. The two test here (see PR header) are failing with a timeout (expected, but not helpful) -> even 2 mins wait for the pyramids is not enough when the whole suite is being run (it is fine if these 2 tests are run singly).
2. There are now 6 regressing tests, ie. 3 and 3 new failing tests are coming for every run of the Trigger. Now we are back at 12 failing tests, because of the 6 "fresh" ones. The new failures, at fist look, have again to do with writing data on disk (pyramids, imports).
3. I am excluding this PR for tomorrows build in the hope that it is this PR's fault (the new failures).
4. The testintegration is now at ca 6h lenght (it was about 5.30 minutes when I started fixing the tests (usually by waiting for pyramids or until a thumb is generated on disk or similar).

cc @jburel