Hi Nate,

PipelineJobs should all be serializable. This is important because many jobs can be executed on different machines and we need a way to transfer them around. Additionally, they may have to sit in a queue for a long time and you don't want them to hold onto a lot of in-memory data. In your particular case, it sounds like your job will always run on the web server, but it still needs to be serializable.

What this means is that you should pass a job the information it needs to look up data, instead of the TableInfo itself. For example, you might pass it a list of run ids, a URL that includes the filters and sorts, etc. Then, inside your job when it starts running, go ahead and create the TableInfo. You may need to create a different MS2Schema.createSpectraCountTable() method that doesn't rely on the ViewContext directly, since that class won't be serializable.

Hope this helps.

Thanks,
Josh