Hi all!
I've recently setup TrueNAS on a X9DRW-3F board and enjoyed everything, except the performance.
I have a all flash setup that is reasonably fast, but definitely saturates the Gigabit network, and even 4Gbps via LAG.
I could simply get a network card, but I saw that the X9DRW-CTF31 board comes with 2 built in 10Gbps ports (RJ45) for 20Gbps total (will probably stick with 10 since I want non-local network connectivity as well.)
The board was also a nice upgrade as I could put in a SAS aom that lets me do more than 4 drives RAID10 at a time.
Here's where the problems began:
I tested with both cpus installed and all memory modules, expecting the board to fire right up.
It did not.
I tested with a single v1 LGA 2011 processor, and to my surprise, it booted!
Supermicro only lets you try one processor at a time on cpu socket 0, so I couldn't try a single cpu in socket 1.
I tested with both downgraded cpus to v1 LGA 2011 processors, and it did not fire up.
I tested wtih one LGA 2011 v2 cpu in socket 0, and booted!
So clearly, there is either something wrong with QPI, or something wrong with socket 1. As far as I can tell, there are no pin damages in socket 1 at all.
I am not getting a no post issue, I am getting quite literally a NO POWER issue.
This should mean that there is a short somewhere, but there is not as far as I can tell. The board is perfectly functional every other way other than that non functioning socket.
Has anyone ever run into this? Any ideas? The latest bios for the board is (yes) from 2015, compared to the x9drw-3f which had a latest bios release at 2019.
The bios is at this latest version.
I've recently setup TrueNAS on a X9DRW-3F board and enjoyed everything, except the performance.
I have a all flash setup that is reasonably fast, but definitely saturates the Gigabit network, and even 4Gbps via LAG.
I could simply get a network card, but I saw that the X9DRW-CTF31 board comes with 2 built in 10Gbps ports (RJ45) for 20Gbps total (will probably stick with 10 since I want non-local network connectivity as well.)
The board was also a nice upgrade as I could put in a SAS aom that lets me do more than 4 drives RAID10 at a time.
Here's where the problems began:
I tested with both cpus installed and all memory modules, expecting the board to fire right up.
It did not.
I tested with a single v1 LGA 2011 processor, and to my surprise, it booted!
Supermicro only lets you try one processor at a time on cpu socket 0, so I couldn't try a single cpu in socket 1.
I tested with both downgraded cpus to v1 LGA 2011 processors, and it did not fire up.
I tested wtih one LGA 2011 v2 cpu in socket 0, and booted!
So clearly, there is either something wrong with QPI, or something wrong with socket 1. As far as I can tell, there are no pin damages in socket 1 at all.
I am not getting a no post issue, I am getting quite literally a NO POWER issue.
This should mean that there is a short somewhere, but there is not as far as I can tell. The board is perfectly functional every other way other than that non functioning socket.
Has anyone ever run into this? Any ideas? The latest bios for the board is (yes) from 2015, compared to the x9drw-3f which had a latest bios release at 2019.
The bios is at this latest version.
