Multi Client Support

ParaView Support
1

We are looking into using ParaView for a multi user production system.
There is an array of up to 10 datasets, each about 10Gb.
The datasets represent a volume with internal cells representing quality information.

The clients are required to be able to independently view the quality model and execute queries to determine aggregate quality within a region of the volume.

Is this use case reasonable for ParaView?
What is the current status of multi-user support? I noticed a recent post suggesting that multi-user support should be deprecated.

Paul

2

“pvserver --multi-clients” is for collaborative work, a single server for multiple users.
This is the option that may be phased out.

In your case, you just need to run a pvserver everytime a user want to connect, it is very standard and will always work.

3

thanks for the feedback.
So, do we have a separate pvserver instance for each user, sharing the common data source?

The users would not normally modify the common dataset but there will be occasional cases where this is required e.g. import into a region of the dataset from an external source.
Should we provide a remote call interface for user updates of the common dataset or is it ok for individual users to directly modify the common dataset. Does the system handle dataset locking to manage updates from multiple users.

Thanks in advance.
Paul

4

So, do we have a separate pvserver instance for each user, sharing the common data source?

Yes, if you consider your server hard disk to be the “common data source”.

Should we provide a remote call interface for user updates of the common dataset or is it ok for individual users to directly modify the common dataset. Does the system handle dataset locking to manage updates from multiple users.

This is your server filesystem responsability. If a dataset is modified, then the user accessing it in the future will see the modified version. The users already accessing it will not see the modification. It is exactly the same as, let’s say, multiple instance of a word processor running on a single machine.