HDD Temps Abruptly Stopped Reporting

[supersean]

Hi Guys,

I'm kind of at a standstill here, any guidance would be very helpful. Even a compass direction :)

After about 3/4/2020, my disk temps were no longer being reported on the dashboard. I read that allowing the disks to spin down and turn off could be part of the reason (and also that there's no need to do so), so I disabled that setting. I browsed through the logs and couldn't find anything that would be immediately alarming, but I'm not very familiar with FreeNAS yet.

I appreciate any help. I'm happy to post any information that would prove helpful.

Thanks.

 

[Redcoat]

This thread any help? https://www.ixsystems.com/community/threads/no-drive-temps-shown-on-dashboard-11-3.84553/

Useful info: how about posting that covered in Forum Rules (red on masthead) https://www.ixsystems.com/community/threads/forum-rules.45124/

Also - here in the forums "Search" is your friend...

 

[supersean]

yeah, I read that thread and NAS-103898 doesn't really address what I'm seeing..unless it does? I'll add a screenshot for clarity:

 

[supersean]

Oops, I don't see a way to edit my post above -- basically even a place for me to go look for logs would be nice. I'm not knowledgeable of FreeNAS yet so I don't even know for example where the time series data (I assume it's stored as that) for disk temps is stored, and I don't know what's doing the writing, or what's reading it from the disks. If anyone could pass a few pointers about what logs to check I'd be very happy to try to find my own solution.

 

[Redcoat]

OK, so you have the display for the temperatures but its not updating ... (I didn't get that from your first post).

I don't have the answers to your questions.

 

[Samuel Tai]

Where's your system dataset? If it's on the boot zpool, you may be out of space. All the graph data is stored in /var/db/collectd/rrd/localhost.

 

[supersean]

Thanks guys for the advice so far! Samuel, that's a good point... I checked here's what I'm seeing for freenas-boot:

Code:

[root@nas /]# zpool list
NAME           SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP  HEALTH  ALTROOT
freenas-boot   119G  6.42G   113G        -         -      -     5%  1.00x  ONLINE  -

I'm going to poke around some more, I appreciate your guys' input. Just knowing there's collectd here gets me started...

 

[Samuel Tai]

Also, look through /var/log/middlewared.log to see if middlewared flagged anything around the time the reporting stopped.

 

[supersean]

Thank for the quick response... I'm also wondering -- would a hostname change mess up graphing for any reason? I did change the hostname recently...

Below is a snippet of the logs:

Code:

[root@nas /var/log]# grep collectd middlewared.log* | grep "2020/03/23"
middlewared.log.1:[2020/03/23 18:00:35] (DEBUG) EtcService.generate():274 - No new changes for /etc/local/collectd.conf
middlewared.log.1:[2020/03/23 18:00:37] (DEBUG) EtcService.generate():274 - No new changes for /etc/local/collectd.conf
middlewared.log.1:[2020/03/23 18:05:57] (DEBUG) ServiceService._simplecmd():287 - Calling: restart(collectd)
middlewared.log.1:[2020/03/23 18:05:57] (WARNING) ServiceService._system():309 - Command '/usr/sbin/service collectd-daemon onestop ' failed with code 1: b'collectd_daemon not running? (check /var/run/collectd-daemon.pid).\n'
middlewared.log.1:[2020/03/23 18:05:59] (INFO) EtcService.generate_all():283 - Skipping collectd group generation
middlewared.log.1:[2020/03/23 22:03:31] (DEBUG) ServiceService._simplecmd():287 - Calling: restart(collectd)
middlewared.log.1:[2020/03/23 22:03:31] (WARNING) ServiceService._system():309 - Command '/usr/sbin/service collectd-daemon onestop ' failed with code 1: b'collectd_daemon not running? (check /var/run/collectd-daemon.pid).\n'
middlewared.log.1:[2020/03/23 22:03:33] (INFO) EtcService.generate_all():283 - Skipping collectd group generation

And as some food for thought, I checked the temperature graphs today... It looks like all of the disks started showing temperature data 2020/05/05 at 12:30 pacific time -- except for one! To re-iterate from my initial post: The only changes I've done were on the 7th, which (to paraphrse) involved disabling the disks' power-saving settings.

 

[supersean]

To continue from the bottom of my recent post -- the last disk that wasn't showing temperature data DID have power-saving settings turned on. I think this is because when I disabled the settings on the 7th (or was it the 5th? ) I performed a bulk-edit from the WebGUI on all 7 disks. 6 are HDDs, but the disk in question is an SSD.

I went back and checked when I made changes to the disk temp settings ... I'm confident I changed them on the 5th, but you don't see any disk temp data in May in the screenshot I posted above! Here's a screenshot today:

I think the lesson I've learned (again) is to write down all the changes I made when. I appreciate you guys for helping me on this. I double-checked the manual too...

 

[Samuel Tai]

Changing hostnames shouldn't affect collectd, so long as you did it through the GUI. You'll need to check /etc/hosts to see if the new name took effect at the bottom of the file.

Bulk edits have bitten me before. I only edit individual disks now.