Did the limits for RAM on Home edition get changed with SFOS 16.05.2 MR2? I'm only seeing 4GB after update. Still only showing 4GB after rolling back to SFOS 16.05.1 MR-1. I was seeing 6GB usable before upgrade.
Let's dig in and see if we can find something.
Post output by running two commands in the console.
console> system diagnostics show memory
console> system diagnostics show cpu
I see JD MaC has a point here, are you guys experiencing this on a hardware appliance? The above command will fetch the hardware information for us.
Finally, a memory test. Refer: https://community.sophos.com/kb/en-us/125025. Note: It may take 1-2 hour time to complete a full scan of 6 GB. Post the output.
Thanks
sachingurung said:Finally, a memory test. Refer: https://community.sophos.com/kb/en-us/125025. Note: It may take 1-2 hour time to complete a full scan of 6 GB. Post the output.
Ok, is this really necessary? This test the physical memory by running a test pattern. I feel pretty comfortable saying that everybody's memory didn't go bad at exactly the same time.
Mine's brand new and i ran clean across 2 full passes in memtest. I always run memory/stability tests before relying on a new system.
And as i mentioned booting the system to any other linux distro shows 8G of ram. Bios shows 8 G of ram. dmidecode shows 2 x 4G.
My only guess would be something in the kernel changed how it deals with memory. it's basically like i don't even have a second stick of memory.
Was there a kernel update between 16.05-MR1 and MR2? Does it fall back to a previous kernel if you tell it to use MR1 again or is that only switch the userland OS
Maybe the kernel can't map the second dimm due to the ram=6G, but that's only a guess.
Has anyone using a single 8G dimm reported the same issue?
Just in case it's CI323 related mine's brand new barebones from newegg, I installed memory and SSD.
Memory: https://www.newegg.com/Product/Product.aspx?Item=N82E16820191523
SSD: https://www.newegg.com/Product/Product.aspx?Item=N82E16820173150
CI323: https://www.newegg.com/Product/Product.aspx?Item=N82E16856173128
The bios version is below
# dmidecode | grep Version: | head -1
Version: B301P019
Mine's brand new and i ran clean across 2 full passes in memtest. I always run memory/stability tests before relying on a new system.
And as i mentioned booting the system to any other linux distro shows 8G of ram. Bios shows 8 G of ram. dmidecode shows 2 x 4G.
My only guess would be something in the kernel changed how it deals with memory. it's basically like i don't even have a second stick of memory.
Was there a kernel update between 16.05-MR1 and MR2? Does it fall back to a previous kernel if you tell it to use MR1 again or is that only switch the userland OS
Maybe the kernel can't map the second dimm due to the ram=6G, but that's only a guess.
Has anyone using a single 8G dimm reported the same issue?
Just in case it's CI323 related mine's brand new barebones from newegg, I installed memory and SSD.
Memory: https://www.newegg.com/Product/Product.aspx?Item=N82E16820191523
SSD: https://www.newegg.com/Product/Product.aspx?Item=N82E16820173150
CI323: https://www.newegg.com/Product/Product.aspx?Item=N82E16856173128
The bios version is below
# dmidecode | grep Version: | head -1
Version: B301P019
Kevin Brierly said:Has anyone using a single 8G dimm reported the same issue?
Yes ... see previous post here.
Single 8GB and seeing same issue.
I dug thru my syslog.log and grabbed memory and kernel command line across all boots since initial install
It did see 8G before the mem=6G was added to boot options even on the MR-2 kernel.
You can also see my attempt to go back to MR-1 to no avail on getting back 6-8G, but the mem=6G option is there.
The timestamp on my license file is Mar 8th 14:21 if that helps at all
Also i looked up the mem= kernel option. This MIGHT be a physical address space collision.
I personally believe this is directly tied to the addition of the mem=6G on people with certain hardware.
mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory
Amount of memory to be used when the kernel is not able
to see the whole system memory or for test.
[X86] Work as limiting max address. Use together
with memmap= to avoid physical address space collisions.
Without memmap= PCI devices could be placed at addresses
belonging to unused RAM.
Mar 8 20:05:28 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_1139 zd=/dev/sda ro ro quiet console=tty0 console=ttyS0,38400n8 ZD=80 SM=0Mar 8 20:05:28 (none) user.warn kernel: [ 0.000000] Memory: 8075208K/8296108K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 220900K reserved)Mar 8 20:23:45 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_1139 zd=/dev/sda ro ro quiet console=tty0 console=ttyS0,38400n8 ZD=80 SM=0Mar 8 20:23:45 (none) user.warn kernel: [ 0.000000] Memory: 8075208K/8296108K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 220900K reserved)Mar 8 14:29:43 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_1139 zd=/dev/sda ro ro quiet console=tty0 console=ttyS0,38400n8 ZD=80 SM=0Mar 8 14:29:43 (none) user.warn kernel: [ 0.000000] Memory: 8075208K/8296108K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 220900K reserved)Mar 8 14:40:16 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 zd=/dev/sda ro ro quiet console=tty0 console=ttyS0,38400n8 ZD=80 SM=0Mar 8 14:40:16 (none) user.warn kernel: [ 0.000000] Memory: 8075208K/8296108K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 220900K reserved)Mar 8 14:42:55 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 zd=/dev/sda ro ro quiet console=tty0 console=ttyS0,38400n8 ZD=80 SM=0Mar 8 14:42:55 (none) user.warn kernel: [ 0.000000] Memory: 8075208K/8296108K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 220900K reserved)Mar 8 14:44:07 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 console=tty0 console=ttyS0,38400n8 quiet maxcpus=4 mem=6G ZD=80 SM=0Mar 8 14:44:07 (none) user.warn kernel: [ 0.000000] Memory: 3952648K/4101804K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 149156K reserved)Mar 8 17:35:56 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 console=tty0 console=ttyS0,38400n8 quiet maxcpus=4 mem=6G ZD=80 SM=0Mar 8 17:35:56 (none) user.warn kernel: [ 0.000000] Memory: 3952648K/4101804K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 149156K reserved)Mar 8 17:38:43 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 console=tty0 console=ttyS0,38400n8 quiet maxcpus=4 mem=6G ZD=80 SM=0Mar 8 17:38:43 (none) user.warn kernel: [ 0.000000] Memory: 3952648K/4101804K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 149156K reserved)Mar 8 20:02:51 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 console=tty0 console=ttyS0,38400n8 quiet maxcpus=4 mem=6G ZD=80 SM=0Mar 8 20:02:51 (none) user.warn kernel: [ 0.000000] Memory: 3952648K/4101804K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 149156K reserved)Mar 8 20:08:18 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_1139 console=tty0 console=ttyS0,38400n8 quiet maxcpus=4 mem=6G ZD=80 SM=0Mar 8 20:08:18 (none) user.warn kernel: [ 0.000000] Memory: 3952648K/4101804K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 149156K reserved)Mar 8 20:12:43 (none) user.info kernel: [ 0.000000] Command line: auto BOOT_IMAGE=16_05_2160 console=tty0 console=ttyS0,38400n8 quiet maxcpus=4 mem=6G ZD=80 SM=0Mar 8 20:12:44 (none) user.warn kernel: [ 0.000000] Memory: 3952648K/4101804K available (5864K kernel code, 812K rwdata, 2152K rodata, 1336K init, 564K bss, 149156K reserved)
I also think it being an pci device collision could explain why some are seeing different amounts of available memory than others.
It would depends on where the PCI devices map into the system at.