# Conduit Data Source
**URL:** https://discourse.paraview.org/t/conduit-data-source/13770
**Category:** In Situ Support
**Tags:** proposal
**Created:** [January 22, 2024, 4:40pm UTC](https://discourse.paraview.org/t/conduit-data-source/13770 "2024-01-22T16:40:14Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)
#### Post date: [January 22, 2024, 4:40pm UTC](https://discourse.paraview.org/t/conduit-data-source/13770/1 "2024-01-22T16:40:14Z")
`CATALYST_DATA_DUMPS` is an excellent way to debug catalyst instrumentation. When set, catalyst dumps out all conduit datasets to a directory and then one is expected to use `catalyst_replay` to replay the simulation. What I am finding myself do is use replay to run a simple pipeline to save the data as `vtpd` and then open it in paraview to inspect. Instead, if I could just open the directory in which all the conduit data is in ParaView directly, that’d be great!
In otherwords, I want a data-reader that reads the data dumps generated by catalyst for each “execute” set and read those in as timesteps in ParaView itself.
Shouldn’t be too hard to implement, but I’d expect be extremely handy to debug simulation / instrumentations. Thoughts?
---
### Author:  [@Lucas\_Givord](https://discourse.paraview.org/u/Lucas_Givord)
#### Post date: [January 23, 2024, 12:59pm UTC](https://discourse.paraview.org/t/conduit-data-source/13770/2 "2024-01-23T12:59:41Z")
it would be really helpful indeed.
About catalyst tools for debugging, your topic remember me this project : [https://gitlab.kitware.com/paraview/catalyst-player](https://gitlab.kitware.com/paraview/catalyst-player) (even if it’s not really what you propose here)
FYI @nicolas.vuaille @Christos_Tsolakis @Louis_Gombert @Francois_Mazen
---
### Author:  [@utkarsh.ayachit](https://discourse.paraview.org/u/utkarsh.ayachit)
#### Post date: [January 23, 2024, 1:37pm UTC](https://discourse.paraview.org/t/conduit-data-source/13770/3 "2024-01-23T13:37:31Z")
Indeed, `catalyst_replay` and `catalyst_player` handle slightly different use-cases. Both, in my mind, are intended to debug visualization or co-processing side of things i.e. I have a simulation talking to catalyst and now I want to see how I setup the viz to get results that I want. This data source, on the other hand, is more useful when I am instrumenting the simulation itself.