RAM consumption

[pnizet]

Hi EveryOne,
As I didn't have the opportunity to get a VPS and has the installation of wekan on my small and old Synology NAS failed, I installed wekan on my windows laptop.
The aim of this install is to use this tool for myself only and to continuously run the app.

And the fact is that I notice quite a huge RAM consumption, the app when it is on with not much activity, is around 400MB of RAM.
Is it normal ?
Is there anyway way to reduce it ?

Many thanks

PS : Is there somebody who manage to install it on a synology NAS (so old that it doesn't support docker) ?
is there somebody who would like to share a piece of server with me ?

[xet7]

@pnizet

How you did install Wekan to your Windows laptop?

How much RAM is in your laptop?

[xet7]

@pnizet

Does you synology NAS have x64 processor?

[pnizet]

@xet7
I followed the wiki tutorial for windows https://github.com/wekan/wekan/wiki/Install-Wekan-from-source-on-Windows and I get onyl 4GB of RAM

My NAS is not a x64...

[xet7]

@pnizet

On Windows, are you using these compatible versions with Wekan? Because for example Node.js 6.x causes 100% CPU usage.

• Node.js 4.8.1
• NPM 4.6.1
• fibers 1.0.15
• MongoDB 3.2.14

Wekan uses Meteor framework that currently works on x64.
It does not work yet on Raspberry Pi #1053 etc ARM processors, that would need some more research and work.

[pnizet]

I have to correct versions

Do you think that those 400MB are "normal" ?

[xet7]

Yes, I don't know how to reduce memory usage.

[nztqa]

See also https://forums.meteor.com/t/how-to-increase-decrease-node-js-memory-limits-using-meteor/34355/5

[ripclaw]

Was wondering if Wekan would work OK on the UDOO X86 Ultra SBC ?? It is a quad core Pentium N3710 2.56GHz with 8GB RAM.
(I am pretty sure it would work fine in the UDOO Bolt, but it is a long wait)

[xet7]

@ripclaw

8 GB RAM is plenty. UDOO x86 supports 64bit Linux distros. I would recommend using Ubuntu 18.04 64bit or Debian 9 64bit with Wekan Snap.

Wekan does run with 2 GB RAM, although more is of course better. Production systems usually have 4 GB or more.