StackStorm / st2

StackStorm (aka "IFTTT for Ops") is event-driven automation for auto-remediation, incident responses, troubleshooting, deployments, and more for DevOps and SREs. Includes rules engine, workflow, 160 integration packs with 6000+ actions (see https://exchange.stackstorm.org) and ChatOps. Installer at https://docs.stackstorm.com/install/index.html
https://stackstorm.com/
Apache License 2.0
6.04k stars 746 forks source link

Customize columns in WebUI's history pane #2026

Open johanherman opened 8 years ago

johanherman commented 8 years ago

Hi,

After many, many workflows runs have populated the WebUI's history pane, where some flows have failed, and some succeeded, it get kinda hard to tell the different runs apart.

It would be nice if it was possible to customize the number of columns visible in the view, and what data these columns should present.

That means that we e.g. could chose to show the unique workflow execution ID in the table, and e.g. in our specific use cases it would be nice to be able to show one of our input parameters to the workflow (which is, internally for us, a unique identifier for each flow, as it represents data coming from an external system).

/Johan

enykeev commented 8 years ago

Hi Johan,

As for number of columns, you can already do that using drop down menu behind eye icon in the top right corner of the list view.

column picker

It's a little bit more complicated with adding new customizable columns. I'm personally afraid it might clutter already complex enough view and turn quite a lot of effort into too little value at the end.

We have plans to redesign the list view and make it look more like in rules. We're also investigating into more sophisticated solution for searching and filtering executions and planning to support our Trace API.

Another 'cheap' solution for "be able to show one of our input parameters to the workflow" would be to sort param line by position property making your unique ID always appear first. If you think it might help you, we should probably create another issue specifically for this change.

johanherman commented 8 years ago

Sorry for the late reply.

Yes, I know about the possibility to enable/disable the columns in your screenshot. But it is missing some kind of column for me to identify the runs from each other.

We've started to implement the Trace features in our workflows, so to be able to search and filter executions and runs on those would help a lot.

Not exactly sure what you mean by "sort param line by position property making your unique ID always appear first" though.

dzimine commented 8 years ago

Hi Johan,

we were on the path of adding search on Trace-tags to execution list, but it didn't make it to the upcoming v1 release - sparing details, we are thinking of a special trace-based views, and return to designing once the current release behind. Your voice counts for search on trace tag, we may reconsider.

What are the other columns / data you are missing on the list? Is it an ability to search for them, or display in the list, or both?

johanherman commented 8 years ago

We have started using trace tags for our workflows, that way we can group together various executions for logging, review and debugging purposes.

Usually a set of raw data from the sequencing machines is associated with one unique tag name. This raw data might be re-processed a couple of times by the same workflow before we deem the end result to be satisfactory (after which the output data from the workflow can be delived to the customers/researchers).

At the moment we have a small script (https://github.com/arteria-project/arteria-packs/blob/master/scripts/trace_runfolder.py) that could be run with e.g. python scripts/trace_runfolder.py --tag <tag_name> | xargs -n1 st2 execution get which helps us find the relevant traces for a specific tag, and then the associated executions.

Being able to view/filter/search executions based on these things in the WebUI would greatly help.