FreeNAS 11.1 crashes frequently on Ryzen 1700

[gegtor]

Here is my uptime chart as you can see my system is crashing very frequently

Here is my hardware info:
CPU: AMD Ryzen 1700
MB: MSI Gaming Pro B350M
RAM: Generic 2 x 8GB 2400mHz non-ECC
GPU: AMD RX 550
Boot: Kingston SSD 120GB A400
Storage: 2 x WD RED 1TB

Here is my use case:
4 computer connected using AFP and putting light load (only documents)
3 running VM's each 2GB RAM and 1 CPU

How it crashes:
Whole server just hangs leaving no info in Kernel log or anything
Boots up normally without any errors after pushing reset button

/var/log/messages before crash

Code:

Feb  9 08:50:59 SKA-SERVER afpd[95711]: Login by PReifur (AFP3.4)

Feb  9 10:03:54 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 10:18:54 SKA-SERVER afpd[1604]: Login by VAndriyants (AFP3.4)

Feb  9 10:18:56 SKA-SERVER afpd[32854]: Disconnected session terminating

Feb  9 10:18:56 SKA-SERVER afpd[4870]: Terminated disconnected child[32854], client rebooted.

Feb  9 11:04:32 SKA-SERVER afpd[95711]: AFP logout by PReifur

Feb  9 11:04:32 SKA-SERVER afpd[95711]: AFP statistics: 38212.43 KB read, 79134.44 KB written

Feb  9 11:04:32 SKA-SERVER afpd[95711]: done

Feb  9 11:20:02 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 11:33:12 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 13:05:55 SKA-SERVER afpd[1604]: afp_zzz: entering extended sleep

Feb  9 13:17:35 SKA-SERVER daemon[3266]: ==> Newer Consul version available: 1.0.5 (currently running: 1.0.0)

Feb  9 13:26:00 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 13:26:44 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 13:44:04 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 13:44:38 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 14:24:04 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 14:24:59 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 14:26:31 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 14:34:18 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 15:15:55 SKA-SERVER afpd[1604]: read: Operation timed out

Feb  9 15:15:55 SKA-SERVER afpd[1604]: dsi_stream_read: len:-1, Operation timed out

Feb  9 15:15:55 SKA-SERVER afpd[1604]: dsi_disconnect: entering disconnected state

Feb  9 17:06:28 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 17:13:51 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 17:27:29 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 17:37:13 SKA-SERVER afpd[29874]: Login by VAndriyants (AFP3.4)

Feb  9 17:37:16 SKA-SERVER afpd[1604]: Disconnected session terminating

Feb  9 17:37:16 SKA-SERVER afpd[4870]: Terminated disconnected child[1604], client rebooted.

Feb  9 17:41:44 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 17:52:49 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 17:56:00 SKA-SERVER afpd[8992]: afp_zzz: waking up from extended sleep

Feb  9 18:04:43 SKA-SERVER afpd[29874]: read: Connection reset by peer

Feb  9 18:04:43 SKA-SERVER afpd[29874]: dsi_stream_read: len:-1, Connection reset by peer

Feb  9 18:04:43 SKA-SERVER afpd[29874]: dsi_disconnect: entering disconnected state

Feb  9 18:04:58 SKA-SERVER afpd[31752]: Login by VAndriyants (AFP3.4)

Feb  9 18:05:00 SKA-SERVER afpd[29874]: Disconnected session terminating

Feb  9 18:05:00 SKA-SERVER afpd[4870]: Terminated disconnected child[29874], client rebooted.

Feb  9 18:08:58 SKA-SERVER afpd[31752]: afp_alarm: child timed out, entering disconnected state

Feb  9 18:08:58 SKA-SERVER afpd[31752]: dsi_disconnect: entering disconnected state

Feb  9 18:08:58 SKA-SERVER afpd[31752]: dsi_disconnect: entering disconnected state

Feb  9 18:21:06 SKA-SERVER afpd[8992]: afp_zzz: entering extended sleep

Feb  9 23:41:55 SKA-SERVER afpd[57938]: afp_alarm: reconnect timer expired, goodbye

Feb  9 23:41:55 SKA-SERVER afpd[57938]: Disconnected session terminating

 

[wblock]

Please update to FreeNAS 11.1-u1, which has an important bug fix for memory allocation. This might not be related to the Ryzen at all.

 

[gegtor]

Okay i will try that soon

Forgot to add
I tested this machine full load for 24h running windows and AIDA64 so this isn't a hardware problem but something related to FreeNAS.

 

[gegtor]

11.1-U1 Installed succesfully

I will see how it's doing in long run

 

[snorbaard]

I have a similar problem, albeit with an AMD Ryzen 1400. FreeNAS will just seemingly randomly cause the motherboard/CPU to quit and give me the red LED of death (the motherboard is a MSI Tomahawk B350 and the LED error I'm getting is "No CPU").

Can anybody help me find which resources can I dig into to see when and why the machine crashes? I have a remote machine tailling /var/log/messages and it sees nothing out of the ordinary at all, so the crash much happen so suddenly that the connection is lost before an error message is dumped.

 

[gegtor]

This looks like a hardware issue


Wysłane z iPhone za pomocą Tapatalk

 

[snorbaard]

This looks like a hardware issue


Wysłane z iPhone za pomocą Tapatalk

I'd love to confirm that (I have the same suspicion), but I don't know where to start. I'd hope there's some method of collecting logs at crash-time. The parts all have warranties, but I need to know what part is causing the issue, and since I can't make it crash on command (except with the ridiculous RealTek built-in NIC - replicating that is pretty simple) I can' very much walk into the supplier's door and say "it's broken"

 

[diskdiddler]

Did you do a 3 or 4 pass memtest x86 on it and a 12 hour prime95?

 

[snorbaard]

Did you do a 3 or 4 pass memtest x86 on it and a 12 hour prime95?

I removed a PCI SATA controller after I happened to be watching the FreeNAS terminal as a crash occurred and it referred to a timeout on the drive connected to it and the problems magically disappeared /ducksforcover

Clearly my faith in 15-year old tech was overly enthusiastic.

 

[Nick2253]

PCI SATA controller

If you need additional SATA ports, I would get a SAS controller. I'm not surprised that a PCI SATA controller was your issue, as these tend to use less-than-well-tested chipsets, and they often have driver issues (especially with FreeBSD).

 

[diskdiddler]

There's another thread now about general stability issues with Ryzen and maybe even in FreeBSD overall? Worrying, hopefully it behaves for you.