Reporting/Dashboard Widgets are still extremely WIP I find.

Joined
Jul 4, 2015
Messages
39
Thanks
6
#1
Seems Reporting is still very WIP. I made a list of my issues. I did just make this thread for the below line but once I started really playing with it I kind of kicked the hornet's nest.

CPU Usage on Dashboard and under Reporting is completely broken for me. On Dashboard, it just says 0% all the way across and in Reporting it is just a blank box that says "CPU".

On the Dashboard, for CPU temp the graph looks like it starts at 35C and goes to 100C. Which I find weird, it should scale or at least be adjustable. My CPU sits around 30-50C so the scale is way off and the line barely moves at all.

Load Averages starts at 0 and goes to 4 on the Dashboard. NetData only goes to 1 and the Dashboard does not state if this is load1, load5, or load15. When it should be displaying all three as it does in Reporting.

Memory Usage is reporting correctly but the graph on the dashboard just sits at 0 all the way across. In reporting the server knows my max memory (64GB) but the graph only goes up to 50GB when using 48GB of memory.

The bottom of the memory and swap graphs under reporting starts at -1GB. Yes, negative 1GB.

HDD temp on Dashboard can be misleading. It shows A HDD temp in a given pool. It isn't even an average of all HDDs in the pool. You have no way of knowing at a glace if you have a hot HDD in that pool unless it just happens to be the first HDD in the pool. Maybe have the HDD icon be green, yellow, or red to show the temperature at a glace. Otherwise, at a glance, you can't know if you have a hot HDD unless you click on each HDD. Also, the background color of the temp line above the temp scale is too close to the filled color making it hard to read. It also clips with the °C under the temp.

Viewing disks under reporting in a Grid only shows 2 columns with a big gap in between them on a 1440p screen.

Pulling the data in reporting causing the page to lock up for a bit. Changing from a grid to columns also does this. If selecting 20 items per page it takes several minutes before the page is usable. While the page is locked up I notice 1 thread on my i7-6800k @ 4.4Ghz is at 100% usage.

All reporting pages default to Grid and will not remember you selected columns. It also does not remember you changed the items per page

Dashboard time stamps are unnecessarily long. "Sat Apr 13 2019 09:32:50 GMT-0400 (Eastern Daylight Time)" is just about the longest most unnecessary thing you could put under a graph thats length is only 10 min.

Dashboard CPU Usage, CPU Temp, Load Averages, and Memory Usage bottom sections have large amounts of unused space that could be made smaller for taller graphs.

Scrolling with a mouse across Dashboard graphs to get information on a specific time causes 100% usage on a CPU thread on the above-mentioned CPU.

I feel like all of this data should be just pulled straight from NetData since it is built in. Maybe even go as far as to pull the graph itself into the FreeNAS GUI. NetData is superior in every way to what has been implemented in the FreeNAS GUI. I have to give some tough love and just say maybe scrap all reporting in the FreeNAS GUI and tell people if they want reporting to just use NetData.

All my FreeNAS hardware is in the signature.

Edit: Whoops this should be in Bug Reporting Discussion. Can a mod assist, please?
 
Last edited:

Chris Moore

Super Moderator
Moderator
Joined
May 2, 2015
Messages
8,538
Thanks
2,663
#2
Please open a ticket on these issues at the bug tracker because I agree there are many problems with the new GUI which is the reason I have not switched to it.
https://jira.ixsystems.com/secure/RapidBoard.jspa?rapidView=1&projectKey=NAS

It is partitularly bad if you have a lot of hard drives. I manage a server at work that has 124 spinning disks and two NVMe drives. The I tested the new GUI on that system and it was completely unusable with so many drives. The plan for iXsystems it to (at some point) use the new GUI on their TrueNAS product, so I would think they will want it to work better for large installations but the way it is now, I wouldn't have it.
 

Meyers

FreeNAS Experienced
Joined
Nov 16, 2016
Messages
193
Thanks
26
#3
Yes the current reporting is generally unusable. For work I use Check_MK which gets most of what I need done. If I need reporting I just pop into NetData.
 
Top