Automating spark configurationdeployment on multiple spark client machines

Hi,

Any idea on how to have this deployed automatically over a few hundred client machines? We are to lock down the spark 2.8.1 client configuration such that there shall be no other options on the server to log on to.

Thanks in advance.

Spark21

What deployment environment are you using?

Hi @spark21, on my environment i used a GPO to deploy to all computers. First a repackaged the exe to msi (via AppDeploy Repackager Community | ITNinja ).

I think he is asking about the configuration deployment, not the Spark itself. There is no easy way to do this. Other than some scripts which will modify Spark’s config file. But if this is new users, then you can deploy a preconfigured config file with a script (config is in the spark.properties at C:\Users\User\AppData\Roaming\Spark).

Alternatively, if you need to make server (Domain) field locked, you can try Spark 2.8.2 and newest Client Control plugin for Openfire (4.0.0+). It can let you lock down that field and some other settings.

Yes,wroot, that is what we might really need. Problem is, we need to lock down all existing users specially since we are to migrate existing users currently running on 2.8.1 SPARK over to 2.8.3 SPARK.

I saw that the spark.properties before a user gets to login is blank, and that it is getting updated once a user is prompted for credentials. How do we make sure that the config is the same for all users before a 1st-time user logs in with credentials, and that it points to the correct domain?

You can take spark.properties file from another user (which is already setup as is should) and remove login information which looks like

passwordandsomerandomsymbols

It will have server field saved, so your users would only have to put their username and password, if you copy such spark.properties file for them. Though in the past i have noticed that Spark takes a bit longer to start if you have server already preconfigured (maybe it is looking for it and it takes a few seconds longer).