FreeNAS 11 reporting - graph bug?

Pitfrr

Wizard
Joined
Feb 10, 2014
Messages
1,531
Hello,

In an initial search I couldn't find any other related post... so here it is:

I'm playing around with FreeNAS 11.2 (FreeNAS-11.2-RELEASE-U1 - Build Date: Dec 20, 2018 23:41) on an HP DL380 G6 (dual Xeon X5670 and 144GB of RAM).
I set it up using the onboard RAID controller (HP P410i). Since the RAID controller can only be configured in RAID-x, I have a hardware RAID5 volume that I use in FreeNAS as a stripe. I know it's not good but it's only for testing... :smile:

I'm not familiar with version 11.x (I'm still using 9.10) but I noticed a strange behavior in the reporting part of the web interface and I don't know if it's due to my browser/system/set-up (I tried with 2 different browsers including Firefox v64). I though I'll give some feedback here to see if it's only local.


So I go to Reporting, I select Disk.
There I have following devices: cd0; da0 and da1. I select da0 which is the striped (hardware RAID volume) and I select all metrics..

I get this:
1547901416108.png


Note that the disk_io_time graph is empty.

On the first graph (disk_time), I do see the cursor with the vertical bar and it displays the according timestamp (here Time: Sat Jan 19 2019 13:34:10 ....) and legend (read and write data) below the graph .

Now when I go over the disk_ops graph, the vertical bar is displayed but the the below timestamp and the legend aren't:

1547901600321.png




This seems to be because the second graph (disk_io_time) is empty: if in the metrics I deselect disk_io_time, all the graphs are displayed and behave correctly:

1547901682637.png



So nothing very important, only cosmetic and I just wanted to see if this behavior was specific to my system or not.
Thanks for your feedback.
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
You should immediately stop using hardware RAID with FreeNAS. Probably this is not working because it's hardware RAID. The onboard controller for this box is just a nightmare.

Out of curiosity, switch to the "legacy" interface, and view the same/corresponding graph. If you don't see it there, either, then presumably it's just the nightmare situation of hardware RAID causing the problem. If you *DO* see data there, then that's yet another thing that shows up in the old UI and not the new one.
 

Pitfrr

Wizard
Joined
Feb 10, 2014
Messages
1,531
Don't worry, I don't plan to use hardware RAID, it was only to test and the occasion to try out FreeNAS 11. So I don't have a long term use foreseen for this set-up.

Thanks for the tip to switch to the legacy interface. I noticed the switch was possible but I didn't try it. I'll check it out.
 

Pitfrr

Wizard
Joined
Feb 10, 2014
Messages
1,531
Just for information, I switched over to the legacy interface: no issue on the graphs there.

But I can't find the corresponding graph.
The graph was disk io time but hard to guess the units and values (since the graph is empty). But from the manual I understand this "time spent doing I/Os (ms)".

In the legacy I have following graphs:
- Disk I/O
- Disk busy --> would be the closest to disk io time again based on the description of disk io time "You can treat this metric as a device load percentage (Value of 1 sec time spent matches 100% of load). "
- Disk latency
- Disk operations
- Pending I/O requests
- Disk temperature

Anyway, on the legacy interface no graph issue. So I suppose this is kind of a graphical bug due to the unavailability of the data for the dik (due to HW RAID) and impacting the rest of the graphs.
 
Top