Good Cheap Server - HP Proliant Microserver 4 BAY - OWNERS THREAD

hopefully getting one this week, have seen a deal for a shade under 110 after cash back,

reason? my 3 hdds in my HTPC (yes 3*3.5inch in slim case) are helping contributing to extreme temperatures by physically blocking airflow
i should be able to fill all 4 drives with 2*3TB 1*2TB and 1*1.5TB. probably wont run any sort of raid

i have NO clue on networks, as demonstrated by my questions in my other thread
 
Last edited:
KIA said:
Code: 
sudo grep -i sata /var/log/dmesg
[    2.117125] ahci 0000:00:11.0: AHCI 0001.0200 32 slots 6 ports 3 Gbps 0x3f impl SATA mode
[    2.600124] ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[    2.600135] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[    2.600161] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[    2.600174] ata6: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[    2.600189] ata4: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
Click to expand...

no benches then to prove its getting sata2 speeds.
iv tested in both a n36l and n40l and both times a drive which is tested to do over 210 megs a second on a proper sata2 mobo, does no more than 145mb/s on the microservers. and bios boot up shows sata1 mode only too., they were both running in ahci mode thats for sure, and in bios max 3.0gbps mode was set but none of the 2 i had did sata 2 speeds.
faulty batch?
 
Cyber-Mav said:
no benches then to prove its getting sata2 speeds.
iv tested in both a n36l and n40l and both times a drive which is tested to do over 210 megs a second on a proper sata2 mobo, does no more than 145mb/s on the microservers. and bios boot up shows sata1 mode only too., they were both running in ahci mode thats for sure, and in bios max 3.0gbps mode was set but none of the 2 i had did sata 2 speeds.
faulty batch?
Click to expand...

I just did a test with ATTO on my N36L and got 270MB/s read off a 64GB Crucial C300. That seems pretty fine to me. Oh, and this is from the stand-alone SATA port (using the hacked BIOS).
 
steveo said:
sorry i only have plex server running on mine didnt want it sat under our tv (misses flat out said no lol) its feeding 2 Mac Minis which it handles at the same time no problem whats so ever (2 1080p MKVs over power line) you will be ok with the 6450 as it seems to be the chosen card for the Microservers HTPC builds, if your a linux user people are opting for the Nvidia GT520
Click to expand...

As far as I know Plex server only uses the CPU when transcoding. I believe there is hardware decoding on certain Nvidia GPU's on Mac OS from a framework Apple released back in April though.

Eitherway, I run Plex on my N40L running server 2012 and it seems to transcode everything I throw at it fine to my iPad
 
rotor said:
I just did a test with ATTO on my N36L and got 270MB/s read off a 64GB Crucial C300. That seems pretty fine to me. Oh, and this is from the stand-alone SATA port (using the hacked BIOS).
Click to expand...

can ya check in hdtune to see in the info tabwhats listed, can ya upload a pic of it? want to see whats listed in udma mode.
 
in bios i had quick boot disabled so when it boots up it showed the drive modes and status, even though 3.0 is selected in bios udma mode 6 shows up for all the drives and they dont operate faster than sata 1 speeds.
 
Are you on the hacked BIOS Mav or the standard one?

Seen some mentions on AV forum of people installing the hacked one to get faster speeds not sure if that's inly for SSD's though.
 
Cyber-Mav said:
in bios i had quick boot disabled so when it boots up it showed the drive modes and status, even though 3.0 is selected in bios udma mode 6 shows up for all the drives and they dont operate faster than sata 1 speeds.
Click to expand...

Connect an SSD to the system, boot Ubuntu live CD and run the disk benchmark utility.
 
Just made an interesting/irritating discovery!

It seems that in Server 2012 (and reportedly 2008 R2 according to this link) if you install the Hyper-V role it disables hardware acceleration on some dedicated GPU's if you have one installed. Spent a few days trying to work out why my HD 5450 wasn't doing DXVA2 in XBMC or MPC-HC under Server 2012 even though it was fine on a fresh install of W7 in the same machine with the same Catalyst drivers (12.10).

That link mentions the HD 5770 so I wonder if it's just an issue with the HD 5xxx series cards or whether the HD 6xxx cards may be affected too?

Either way, I'm glad it's sorted as it was noticeable not having hardware acceleration in certain films/tv shows that have fast moving scenes when the CPU maxed out at 100%!
 
the mobo chipset on this microserver is very poor in performance, i get over 22% cpu usage doing a harddrive bench test in hdtune, unless this cpu really is that slow compared to my old 2ghz opteron dual core which used to manage around 2% cpu usage.
 
I've found an interesting discovery too.

I've got an N36L Microserver with ESXi 5.1 on it with a 160GB disk and 2x 1TB Samsung F3's and an HP 500GB Disk. On it at the moment I have only one VM running which is the WHS 2011 installation that does my media serving and backup's of my families PC's.

If I do a disk benchmark using Crystaldisk mark I get around 120Mbit/s ish for sequential read/write within the virtual machine.

Now, I've deployed 5 or so N40L Microservers in exactly the same fashion (1 WHS2011 VM), albeit with different disks (usually WD greens). However, with both ESXi 5 and 5.1 I get comparatively terrible disk throughput benchmarks compared to my N36L.

Here's an example:

1st Test, Crystaldiskmark x64 N36L. ESXi 5.1 on USB stick, Target disk is a Samsung F3 1TB, with a 930GB VMFS formatted virtual disk with NTFS formatted disk in VM.

G6UZA


2nd test, Brand new N40L, ESXi 5.1 on USB Stick, Target Disk is a WD green 2TB, with a ~1800GB VMFS formatted virtual disk and NTFS disk in VM.

7rpci


I'm wondering if this is an issue with ESXi 5.1, the performance of the actual HDD's or something else entirely.

Any ideas?

Thanks

CK
 
since your testing different drives in each server the results are not valid.
test the same drive in each server and see the results. also if your testing from within a vm then the vm's location on the drive will effect speed.
also if your using 100mb as the test amount you should be seeing considerably faster scores since caching would make a big difference,
unless you got disk cache disabled.
 
Cyber-Mav said:
since your testing different drives in each server the results are not valid.
test the same drive in each server and see the results. also if your testing from within a vm then the vm's location on the drive will effect speed.
also if your using 100mb as the test amount you should be seeing considerably faster scores since caching would make a big difference,
unless you got disk cache disabled.
Click to expand...

I can't test my disk in the other persons server, nor their disk in mine!

What do you mean by the VM's location on the disk?? The VM's OS drive (the default supplied 250GB disk) has the OS portion on it. The drive I'm testing is added as an additional hard disk (in its entirety) made up of a VMFS virtual disk presented to WHS 2011 and then formatted as NTFS.
 
tbz_ck said:
I can't test my disk in the other persons server, nor their disk in mine!

What do you mean by the VM's location on the disk?? The VM's OS drive (the default supplied 250GB disk) has the OS portion on it. The drive I'm testing is added as an additional hard disk (in its entirety) made up of a VMFS virtual disk presented to WHS 2011 and then formatted as NTFS.
Click to expand...

Think he means that if the VM was physically located towards the center of the platter on the HDD then it would be accessed faster than if it was located towards the outer edge of the platter. If you run any HDD test on a mechanical HDD you'll see the results peak towards the beginning and tail off as the heads move out towards the edge of the platter.
 
cokecan72 said:
If you run any HDD test on a mechanical HDD you'll see the results peak towards the beginning and tail off as the heads move out towards the edge of the platter.
Click to expand...

Not that it matters, but they peak at the edge and tail off as they get nearer to the spindle. There's more room for data nearer the edge and the speed the platter is moving under the heads is higher.
 
bremen1874 said:
Not that it matters, but they peak at the edge and tail off as they get nearer to the spindle. There's more room for data nearer the edge and the speed the platter is moving under the heads is higher.
Click to expand...

lol I had a feeling I got it the wrong way round, thanks for correcting me :)
 
I've noticed an issue in WHS 2011 where when I copy a file/movie over to my videos folder and check it on one of my media extenders (Sony TV, Xbox 360) it isn't there. For some reason it isn't updating straight away!

When I check it on my PC/laptop it's there and I can play it fine.

Any idea what permissions/account the media extenders use or what service/application is used to update the folders to make them visible on the media extenders?
 
You must log in or register to reply here.
Back
Top Bottom