So my temporary home office is next to the furnace room. Today I started to hear disk thrashing on my new TNAS and found it odd as the disks are brand new. I have also noticed random reboots for no good reason that seem to happen every day. The SSH and web interface were also very sluggish. I finally decided to take a look via SSH. I also configured SYSLOG to send all logs to my syslog-ng docker instance. The disks were powering down (standby) and immediately spinning up every few seconds.
SYSLOG-NG showed:
Jul 4 09:44:59 TNAS-F2-212 kernel: [ 4868.371742] ata1.00: exception Emask 0x0 SAct 0x4 SErr 0x0 action 0x6
Jul 4 09:44:59 TNAS-F2-212 kernel: [ 4868.373349] ata1.00: waking up from sleep
Jul 4 09:44:59 TNAS-F2-212 kernel: [ 4868.374351] ata1: hard resetting link
Jul 4 09:44:59 TNAS-F2-212 kernel: [ 4868.851737] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Jul 4 09:44:59 TNAS-F2-212 kernel: [ 4868.854405] ata1.00: configured for UDMA/133
Jul 4 09:44:59 TNAS-F2-212 kernel: [ 4868.855491] ata1: EH complete
and it did so for both disks every second or so. Sigh. I got the NAS to avoid the issues of using a USB drive and now this!
After a LOT of looking, following google searches that didn't help, disabling standby with HDPARM and SMARTCTL, etc, I did a TOP and noticed something odd. When I was hearing the thrashing /usr/sbin/ter_smartfan was running and it would spawn hdparm -Y /dev/sda and sdb. So I ran /usr/sbin/ter_smartfan manually and sure enough every second it did HDPARM -Y. So I did a ps for ter_smartfan and killed it. Now there is no thrashing or resetting the link. SSH and web interface are snappy. Logging is not showing issues.
If I run /usr/sbin/ter_smartfan manually the issues start again. So I plan to NOT do that. I will keep an eye on things and see if I have future problems or if the random reboots stop.
Just thought I'd share for those experiencing the same issue. Perhaps I'm off base with my reasoning or solution, If so please inform me.
TNAS F2 212 Disk Thrashing
Re: TNAS F2 212 Disk Thrashing
Thank you for your report! We will verify the issue you reported.
To contact our team, please send email to following addresses, remember to replace (at) with @:
Support team: support(at)terra-master.com (for technical support only)
Service team: service(at)terra-master.com (for purchasing, return, replacement, RMA service)
Support team: support(at)terra-master.com (for technical support only)
Service team: service(at)terra-master.com (for purchasing, return, replacement, RMA service)
Re: TNAS F2 212 Disk Thrashing
In order to find out the problem you encountered, we can arrange for you to have an engineer to investigate the problem remotely.
If you agree, please contact our technical support mailbox or the customer service window on the homepage of our website to schedule an appointment. Please remember to provide them with a link to this post.
If you agree, please contact our technical support mailbox or the customer service window on the homepage of our website to schedule an appointment. Please remember to provide them with a link to this post.
To contact our team, please send email to following addresses, remember to replace (at) with @:
Technical team: support(at)terra-master.com (for technical support)
Service team: service(at)terra-master.com (for purchasing, return, replacement, RMA service)
Technical team: support(at)terra-master.com (for technical support)
Service team: service(at)terra-master.com (for purchasing, return, replacement, RMA service)
Re: TNAS F2 212 Disk Thrashing
Yes, please check it out. It's been stable for me since with zero thrashing. All notifications as shown in my initial log snippet have stopped cold. I have also experienced NO (yes capitol for emphasis) random reboots. This was happening most every day before. I found many threads on this site and others complaining about the random reboots with no clear solution. For me, cancelling the ter_smartfan process fixed ALL (again for emphasis, not shouting) my TNAS issues. I expect there is something wrong with that binary or it calls an external command improperly. It could also be an issue with HDPARM itself when ter_smartfan calls it. One thread I found indicated that HDPARM -Y should probe but not wake up the disks if asleep but it does. Running it with a lowercase "y" didn't seem to wake them up and cause the issue.
I don't think my configuration is non-standard. The only thing I modified myself is having syslog send logs to my syslog-ng server. I also only use SSH and NFS, not CIFS, etc. I do believe support or dev teams should check out the ter_smartfan file.
For TNAS users experiencing random reboots I suggest as a test you kill the ter_smartfan process and see if your problems go away.
Re: TNAS F2 212 Disk Thrashing
No offence, but that's not going to happen. I am more than experienced in sysadmin and engineering work with MS, Linux, HP, CISCO, and NETAPP. Not that I think I know it all, but I can work on my own stuff thanks. I just wanted to bring this to the attention of the TerraMaster team and let them consider the issue. If not repeatable it will hopefully go away after the next TOS update.