jmeter-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Philippe Mouawad <>
Subject Re: MongoDB driver
Date Fri, 04 Oct 2019 17:05:25 GMT
On Friday, October 4, 2019, Vladimir Sitnikov <>

> >I propose to drop it for this release 5.2
> This is interesting.
> Do we need Bolt then?

This has been contributed by members of a company tightly related to Neo4j.
MongoDB was not.
I expected MongoDB to be supported by the company

It brings people to JMeter and Graph a field that is expanding.
We try and when something is not successful we drop it, that’s ok for me.

> Who will maintain it?

We’ll see .
I expect initial contributors, I’ll also try to

> >Possible benefits for users is cleaner UI menus
> What if we add a "import-like" feature to the thread group?
> In other words, what if ThreadGroup (or it's ImportConfigurator) includes a
> set of checkboxes:
> "I want Mongo in this threadgroup"
> "I want JSR223 in this threadgroup"
> Then context menus might show only "already imported" elements, to it would
> be very small.
> "not yet imported" could be either completely hidden, or put into
> "unimported yet" sub-menu.
> The user could configure different imports for different thread groups.
> Menus are small, and everybody's happy.

let’s make a POC of it.
But even with it, we should drop MongoDB

> As a bonus, we could even add/remove UI elements from ThreadGroup (or
> whatever element) when the set of imports change.
> For instance, if the user imports "http testing", then we could show "clear
> caches on each iteration" feature to the ThreadGroup / LoopController
> configuration.
> mongo-java-driver-2.11.3.jar is 400KiB.
> It is 10 times smaller than Saxon-HE-9.9.1-5.jar which is ~5.5M
> It is 4 times smaller than commons-math3-3.6.1.jar which is 2.2M
> and so on.
> We could try reducing the initial download size by downloading dependencies
> from MavenCentral-like repositories, however, that is a different story.
> The removal of Mongo does not reduce the download.

It’s not about size, it’s about quality and « «uptodateness »
And also about number of elements but your proposal would help

> Vladimir

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message