On 02/01/2018 09:03 AM, Adam Armstrong wrote:
If anyone has any decent working code to do this we'd consider incorporating it into the code
Issues I can see are :
Handling new/different metrics between device types (HDD vs SSD?) without needing to change agent code (functionally impossible once code has been distributed) or RRD formats (possible, but a pain). Presenting the information in a useful/usable manner (generic handling is good for easy alerting, bespoke handling are good for interactive usability)
Forget HDD vs SSD, different brands of HDD will present functionally equivalent data using different names/attirbute ids, plus the issue of whether the raw or processed data is more useful, and that's after smartmontools has done it's best to clean it up.
Most of my unix agent code is trying to map this into a consistent set of values across the few brands of drive I have.
Handling SMART data seems like it could well be something that's worth handling in a bespoke manner, perhaps with a "fake" status element exposed to the alerting system based on sane thresholds (does the device supply these?).
Every attribute has a processed value which is supposed to run from 100 - 0, where 100 is healthy and 0 is boned. In theory, anyway. But for some of them, you just have to watch it and see if it's trending in the wrong direction.
I'm more of a network person than a systems person, so I'm not as familiar with the metrics available as some might be.
I coded this up mostly because I wanted to keep an eye on my SSD wearout so I could pro-actively replace them when the reach EOL, and collected some other stuff 'cause I could. Of course later I've read that you can't even rely on that!
adam.
On 2018-02-01 13:10:16, Simon Mousey Smith simonsmith5521@gmail.com wrote:
Id be happy to try the code too,
anything to help the community and the development of observium!
Simon
On 31 Jan 2018, at 20:34, Tony Lill wrote:
I didn't use that as a base, mostly because the links to the source code seem to be dead, but I've rolled my own with the latest CE. Wouldn't mind another set of eyes on it. I'll e-mail you, and anyone else interested, a tar file and instructions off-list.
On 01/31/2018 12:38 PM, Attila Nagy wrote:
Hi everyone!
I recently stumbled upon a blog post about making the Unix Agent deliver disk SMART values as well. The only problem is the post is from 2014 and as I saw there has been quite some changes in Observium since that time.
Is there anyone who managed to make this (or any other) solution to
work?
Is there some plan to integrate SMART monitoring into Observium?
Regards, Tylla Ps: If I remember correctly there was some discussion about this a few years back but there wasn't any final result.
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
-- Tony Lill, OCT, ajlill@AJLC.Waterloo.ON.CA President, A. J. Lill Consultants (519) 650 0660 539 Grand Valley Dr., Cambridge, Ont. N3H 2S2 (519) 241 2461 --------------- http://www.ajlc.waterloo.on.ca/ ----------------
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium