Behavior in case of CLARK error

Precondition:

1. A test computer / virtual machine has 5 Gb RAM.
2. Metagenomics framework is set up (at least, CLARK-l databases).

Scenario:

1. Create a workflow with element "Classify Sequences with CLARK" (or use a workflow from the test server with the corresponding test data).
2. Run the workflow with the "bacterial and viral" reference database.

Current state:

• CLARK fails because there is not enough RAM.
• There are no errors in the UGENE log.
• There is CLARK output file on the dashboard.
• There is no error messages from "Classify Sequences with CLARK" element.
• For the "Classification Report" element, connected to the CLARK element there are two messages on the dashboard. The first one is an error message:

  Cannot open classification report: file_name
  

  The second one is an info message:

  There were 0 input reads, 0 reads were classified.
  

Expected state:

• CLARK fails because there is not enough RAM.
• There is an error message in the UGENE log.
• There is no CLARK output file on the dashboard.
• CLARK has a meaningful error message on the dashboard.
• The CLARK element doesn't send the result to the output port, further elements in the workflow, connected to this are not executed. Thus, there is no output from "Classification Report", connected to CLARK.
• There is an output from other not connected elements!