Selecting cluster in the secondary analysis wizards

[Nicklas Nordborg]

A new cluster is coming soon, and it will be based on Slurm and run jobs within Singularity containers (see #1374 and #1375).

It is most likely that not all steps in the analysis pipeline will be available on the new cluster from the start. This means that when starting an analysis step the wizard should only display the clusters that include configuration settings for that step.

This also affects the auto-confirm functionality that currently start the next step on the same cluster as the previous step used. This need to be changed so that a different cluster is selected if the next step is not available on the same cluster.

[Nicklas Nordborg]

In 6632:

References #1376: Selecting cluster in the secondary analysis wizards

The servlet has been updated to allow filtering on more information. By default, it will no longer include clusters that are "not connected" unless the includeNotConnected flag is included in the request. This flag should only be used on the admin page.

If the config parameter is included in the request the servlet will only return clusters that has that configuration option. This has been tested with the Start demux wizard which require that the <demux> section is present for the cluster.

[Nicklas Nordborg]

In 6633:

References #1376: Selecting cluster in the secondary analysis wizards

Most secondary analysis wizards have been updated.

[Nicklas Nordborg]

In 6635:

References #1376: Selecting cluster in the secondary analysis wizards

Implemented a method for automatically selecting a different cluster if the current one is not configured for use with a particular analysis step.

Implemented in the auto-confirm handlers for demux and sequencing run.

[Nicklas Nordborg]

In 6637:

Fixes #1376: Selecting cluster in the secondary analysis wizards

All auto-confirmation handlers should now be updated.