Server opinion

TPerry

Well-known member
Thinking about getting another server nearer to my home location. Planning on having ProxMox installed on it and it replacing a dual Xeon 5520 24GB 2 1TB software RAID 1 system I have (will be about $90 more).

This look like a good deal?

Intel Xeon E3 1240 V3
32GB DDR3 RAM
6x240GB SSD
LSI-9271 w/ CacheVault
10TB @ 1gbit port
/26 IPv4
IPMI
$189/month
 
Last edited:
I have had excellent luck with E3's. They will absolutely blow the pants off of your old 5520's.
 
Server specs look good but it looks like a plan you would get from Getdedi.com ? But I'd be concerned about where their servers are colocated at.

GetDedi's Dallas datacenter is through Incero I believe and if you search WHT forums you will see how they have jacked up their prices without any notice to customers by up to 70% due to Zayo's acquistion of CoreXchange http://www.webhostingtalk.com/showthread.php?t=1355017 & http://www.webhostingtalk.com/showthread.php?t=1355050. Without understanding how that effect's GetDedi's pricing it could be risky if there's any other price rises along the chain which GetDedi may need to pass on to you. And as you can see from the Incero thread, monthly pricing is no guarantee that next month is the same price. Only guarantee is to to pay for more months in a advance which can be risky in itself on a new never used before web host.

GetDedi's Buffalo, New York datacenter is through Colocrossing who have a very bad rep if you search WHT, lowendtalk.com, vpsboard.com
 
Seems nice, where from?
Company called getdedi.com (they have two server centers, Dallas, TX and Buffalo, NY.
Since I'm wanting Dallas I am going to have to go with an E3 1230.

Server specs look good but it looks like a plan you would get from Getdedi.com ? But I'd be concerned about where their servers are colocated at.

GetDedi's Dallas datacenter is through Incero I believe and if you search WHT forums you will see how they have jacked up their prices without any notice to customers by up to 70% due to Zayo's acquistion of CoreXchange http://www.webhostingtalk.com/showthread.php?t=1355017 & http://www.webhostingtalk.com/showthread.php?t=1355050.

GetDedi's Buffalo, New York datacenter is through Colocrossing who have a very bad rep if you search WHT, lowendtalk.com, vpsboard.com

Nice thing about if they do jack it up... I can also say bye-bye. It's a simple matter of backing up the VPS with ProxMox and then just importing it into another server. :p
I Googled for reviews, but didn't find much. I went ahead and ordered one to try out.
 
Yeah i guess if you have servers elsewhere, moving is much easier than folks who have one sole server/location :)

Yeah GetDedi and it's parent company are fine, it's just further upstream needing more research.

Good luck !
 
Yeah i guess if you have servers elsewhere, moving is much easier than folks who have one sole server/location :)
Yep, this one will be replacing a ProxMox at ServerComplete. All the sites I am running on it can either be moved to a new ProxMox server elsewhere easily or I can just move the sites they are hosting (the VPS's) to my other server at SC. I think I may contact SC and see if they have anything comparable (don't think they do though as they apparently purchase older servers and then put them into place and then lease them out).

Yeah GetDedi and it's parent company are fine, it's just further upstream needing more research.
Good luck !
That (whether GetDedi was reliable/good) was one of my main concerns. I'm sure they have a longer term contract with their provider(s). But again, if the prices increase then I can always go elsewhere. It doesn't take me but about an 1.5 hours now to set my servers up since I've done it so many times already. :D
 
Hardware wise, so far I can't say I'm sorry about the server
Code:
root@orion:/home/tracy/work# dd if=/dev/zero of=/tmp/output bs=8k count=10k; rm -f /tmp/output
10240+0 records in
10240+0 records out
83886080 bytes (84 MB) copied, 0.067347 s, 1.2 GB/s
root@orion:/home/tracy/work# dd if=/dev/zero of=test bs=1048576 count=2048
2048+0 records in
2048+0 records out
2147483648 bytes (2.1 GB) copied, 1.75423 s, 1.2 GB/s
root@orion:/home/tracy/work# dd if=test of=/dev/null bs=1048576
2048+0 records in
2048+0 records out
2147483648 bytes (2.1 GB) copied, 0.243078 s, 8.8 GB/s

and a test file pull from another post wanting folks to check his network speed

Code:
wget http://67.215.237.122/100.MB
--2014-03-18 09:32:20--  http://67.215.237.122/100.MB
Connecting to 67.215.237.122:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/mathematica]
Saving to: `100.MB'

100%[===================================================================================================================================================>] 104,857,600 10.9M/s   in 9.4s   

2014-03-18 09:32:30 (10.6 MB/s) - `100.MB' saved [104857600/104857600]
 
nice... how about ioping numbers https://code.google.com/p/ioping/ ?
Any particular call to it you want to see?

Code:
root@orion:/home/tracy/work/ioping-0.8# ./ioping -c 5 .
4.0 KiB from . (ext3 /dev/mapper/pve-root): request=1 time=58 us
4.0 KiB from . (ext3 /dev/mapper/pve-root): request=2 time=78 us
4.0 KiB from . (ext3 /dev/mapper/pve-root): request=3 time=59 us
4.0 KiB from . (ext3 /dev/mapper/pve-root): request=4 time=61 us
4.0 KiB from . (ext3 /dev/mapper/pve-root): request=5 time=76 us

--- . (ext3 /dev/mapper/pve-root) ioping statistics ---
5 requests completed in 4.0 s, 15.1 k iops, 58.8 MiB/s
min/avg/max/mdev = 58 us / 66 us / 78 us / 8 us

Code:
root@orion:/home/tracy/work/ioping-0.8# ./ioping -R /dev/sda

--- /dev/sda (device 1.8 TiB) ioping statistics ---
314 requests completed in 3.0 s, 104 iops, 419.6 KiB/s
min/avg/max/mdev = 1.3 ms / 9.5 ms / 16.8 ms / 2.9 ms
root@orion:/home/tracy/work/ioping-0.8# ./ioping -RL /dev/sda

--- /dev/sda (device 1.8 TiB) ioping statistics ---
3.9 k requests completed in 3.0 s, 1.4 k iops, 338.1 MiB/s
min/avg/max/mdev = 144 us / 739 us / 17.0 ms / 1.5 ms

Code:
root@orion:/home/tracy/work/ioping-0.8# df
Filesystem            1K-blocks    Used  Available Use% Mounted on
udev                      10240       0      10240   0% /dev
tmpfs                   3287928     360    3287568   1% /run
/dev/mapper/pve-root   99083868 4354676   89696028   5% /
tmpfs                      5120       0       5120   0% /run/lock
tmpfs                   6575840   18648    6557192   1% /run/shm
/dev/mapper/pve-data 1773703684  200880 1773502804   1% /var/lib/vz
/dev/sda2                505764   35458     444194   8% /boot
/dev/fuse                 30720      12      30708   1% /etc/pve
root@orion:/home/tracy/work/ioping-0.8# free
             total       used       free     shared    buffers     cached
Mem:      32879264    5209436   27669828          0    1040684    3566992
-/+ buffers/cache:     601760   32277504
Swap:     32505848          0   32505848
 
Currently SFTP'd in and transferring the setups from my old ProxMox box over and then will import them into the new one as well as getting ready to configure CSF on it.
 
nice ioping numbers, only one out of place is the seek rate for /ioping -R /dev/sda ~9.5ms avg seems very slow compared to rest of the numbers

what brand/model SSD?
 
This look a little better?

Code:
root@orion:/home/tracy/work/ioping-0.8# ./ioping -R /dev/sda1

--- /dev/sda1 (device 1.0 MiB) ioping statistics ---
88.5 k requests completed in 3.0 s, 30.4 k iops, 118.6 MiB/s
min/avg/max/mdev = 23 us / 32 us / 1.3 ms / 27 us
root@orion:/home/tracy/work/ioping-0.8# ./ioping -RL /dev/sda1

--- /dev/sda1 (device 1.0 MiB) ioping statistics ---
17.9 k requests completed in 3.0 s, 6.6 k iops, 1.6 GiB/s
min/avg/max/mdev = 139 us / 151 us / 7.8 ms / 59 us

Code:
Disk /dev/sda: 1999.3 GB, 1999307276288 bytes
255 heads, 63 sectors/track, 243068 cylinders, total 3904897024 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disk identifier: 0x00000000

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               1  3904897023  1952448511+  ee  GPT
Partition 1 does not start on physical sector boundary.

Disk /dev/mapper/pve-root: 103.1 GB, 103079215104 bytes
255 heads, 63 sectors/track, 12532 cylinders, total 201326592 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disk identifier: 0x00000000

Disk /dev/mapper/pve-root doesn't contain a valid partition table

Disk /dev/mapper/pve-swap: 33.3 GB, 33285996544 bytes
255 heads, 63 sectors/track, 4046 cylinders, total 65011712 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disk identifier: 0x00000000

Disk /dev/mapper/pve-swap doesn't contain a valid partition table

Disk /dev/mapper/pve-data: 1845.2 GB, 1845225324544 bytes
255 heads, 63 sectors/track, 224335 cylinders, total 3603955712 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disk identifier: 0x00000000

Disk /dev/mapper/pve-data doesn't contain a valid partition table
 
definitely more inline with the hardware there :D

raid 10 6x ssd right ? could be Intel 530 SSDs, they on special now at $150-180 for 240GB versions
 
And now the fun begins. Uploading the associated images of the OS's I want to be able to use, then moving several sites over from a dedicated server to custom (made by me! :p) VPS's I will have to configure. I'm still going to keep at least one of my ServerComplete servers online.
 
Dang - first site didn't take long at all (was a copy from my existing ProxMox server). Hardest thing was reconfiguring the network. Simple WordPress site with Trellis Desk on it. And it DOES seem to be faster than the ProxMox server I had it on (which was using software RAID 1 with SATA 3 1GB drives).
 
Use fio to get real disk stats.

You can check, no need for pretty sure. :)
After being up about 30 hours... I was too tired to check. :)
And actually... no I can't with that command ;)
Code:
root@orion:/home/tracy# hdparm -i /dev/sda | grep Model
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0d 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
HDIO_GET_IDENTITY failed: Invalid argument
 
Top Bottom