Hard Drive Monitor and health check

Suggestions & Feedback
#1

On the dashboard we have CPU/RAM usage and do have a hard drive usage for how much you have used. But I am looking for disk utilization. I feel would be good to know sometimes how hard the disk are getting hit. I’m very much not a linux person and more windows so I know tools for windows, both built in and that you can download to see that and also monitor the health of the drives. Is something like that possible with linux, hopefully not by command line. Would be great if part of the dashboard. My box has 4x3TB drives and I really have no idea how good or possibly bad they are being used.

ps. if this is better of not being a suggestion and just something for general chat that it’s ok to move.

#2

my personal favorite system monitor is netdata…its incredible and very light weight: http://my-netdata.io/ and now does alerting. You could embed custom sections in the QB panel.

There is a request for a SMART monitoring plugin for it.

1 Like
#3

I’ve have definitely considered this as a feature. The largest thing preventing me from jumping on this idea fully is the large majority of hardware raid controllers. Things would be easier if smartctl used the same parameters hard or soft, but hardware raid makes one virtual drive, whereas software raid leaves the drives exposed.

With SWR the call to smart status is:

smartctl -a /dev/sda
smartctl -a /dev/sdb

With HWR the call to smart status first requires a call to the raid controller to get the device IDs then a specific call to smartctl for the raid controller + ID:

smartctl -d megaraid,4 /dev/sda
smartctl -d megaraid,5 /dev/sda

or

smartctl -d 3ware,1 /dev/tw1
smartctl -d 3ware,2 /dev/tw1

The list goes on.

Further, I’m not sure how people would react to giving apache smartctl privileges. Feel free to discuss :slight_smile:

#4

you have a point. If you look at the main players in the DIY NAS platforms, you need your raid controllers flashed to IT mode which will disable the raid functions to present the smart status for each disk. Probably the only way you could support this is with software raid.

#5

Not to highjack the thread, but I think this fits inside the “hard drive monitor” part as well, since we are speaking of RAIDs. With my setup I have a RAID0 (two 36GB) for my OS/apps drive, and then another RAID0 (two 500GB) for my actual torrents.

Anyway to have the dashboard be able to show information for all hard drives attached to the system?