03-05-2016, 10:13 PM
Introduction
VPS 5 is hosted by https://cobalthost.net/ and was given to me during the last VPS giveaway. It's located in Canada and uses VMWare as visualization technology, which offers great performance and flexibility. Specs seem fine, especially 50GB diskspace and unlimited traffic.
Specs in detail:
The VPS has 2 Xeon Cores
Xeon looks awesome on first glance, but if you look it up, it is a 9 year old CPU and that shows, as we will see in later benchmarks.
http://ark.intel.com/products/28443/Inte...33-MHz-FSB
Installation
cobalthost is using their own Panel, which fulfills the basic needs, shows cpu/ram/hdd usage and offers the usual suspects like reboot/reinstall etc.
CPU Usage is always wrong in the panel, mobil version does not work. Installation works, which is most important.
One can choose to install on DC-2 or DC-3, since nowhere i can see the utilization of either node or see what hardware is in it it's is a gamble.
I first chose DC-3, which was a mistake. It is so slow that there can be only 2 reasons for it. a) it has like hundreds of VMs running, or b) the hw is dying (as in faulty), possible the hdds.
Booting a VM takes up to 5 minutes, installing ~30 minutes.
So i reinstalled on DC-2, which works better, boot and install times are normal. Installation went without issues. Their install script leaves a /bin/npasswd script back and a non working rc.local entry in case of debian. On FreeBSD i could not find any artifacts but 2 installed packages.
There is a slight installation bug with FreeBSD, which i already reported. (swap missing), also it installs on UFS2 instead of ZFS, which is a bummer but should do for most tasks.
One cool with with cobalthost is that you can have as many VPSs as you have resources for, so with 2 cores you can do 2 VPSs. Very cool.
Benchmark
Since i am running FreeBSD the usually FreeVPS benchmark script won't work, so i had to improvise and used unixbench.
Unixbench:
Now that looks kind of horrible. As a small comparison a FreeBSD VM on an older core i3 under load in production had a Final Score of "231". FreeBSD VM in an older Xeon also in production with load on it did "362".
ioping:
As you can see, it is worse than on a raspberry pi on sdcards, the 300ms+ peaks are especially bad.
dd:
Now that is also right up the raspberry pis disks speed as well. In all fairness, doing the test during different hour yield better results at around 12-30MB/s
Summary: Disk speeds and latency are bad.
CPU:
I tested CPU speed with openssl speed which every sysem has for comparison:
That should be enough to get a rough idea and be able to compare with his/her own VPS/Desktop
Network:
Now the specs say 100mbit, but in reality it is connected with gbit, which is nice. Speeds inside Canada are good, ok to the USA and not so good to Europe but that was to be expected.
upload speeds are worse, why that may be, i think everyone can guess. Generally the speeds are good enough for any personal use, for me at least its enough.
My VPN does 15mbit (compressed) to my home link, which is enough for my VPN related activities. Also enough for mumble/teamspeak or a small private website.
VPS Usage
I mainly use it for playing around with Operatingsystems and run a mumble and an OpenVPN server, both private for myself. Maybe i'll use to backup some stuff, since the 50GB disk space is nice.
Reliability
Can not say much about that yet, for now it has been stable, if it ever goes down i will update this thread.
Conclussion
Well that is a mixed bag, it is very old, dying hardware and probably over sold into oblivion. So if you are looking for a VPS where you actually have to run something that needs power or disk speeds from this century you are out of luck. A quad core CPU for half their customers(DC-3 has 6 cores i think) is simply not enough, also the apparent lack of a storage for the VMs other than the included HDDs on the old server shows, very much so. If you are looking for a VPS to do some private stuff, test your code, learn linux, run some small daemons or anything similar like that it does its job, after all it's free and for free it's good, can't really complain. For anyone looking to invest their own money i would suggest to look for another hoster.
greetings
p.s. redid all test with debian, same result, does not seem to be OS related.
VPS 5 is hosted by https://cobalthost.net/ and was given to me during the last VPS giveaway. It's located in Canada and uses VMWare as visualization technology, which offers great performance and flexibility. Specs seem fine, especially 50GB diskspace and unlimited traffic.
Specs in detail:
- Disk Space 50 GB
- RAM: 2 GB RAM
- IP Addresses: 1x IPv4
- Virtualization: VMWare
- Monthly Traffic: Unlimited
- Location:  Canada
- Control Panel: Cobalt Panel
- Connection: 100 MBit/s
The VPS has 2 Xeon Cores
Code: (Select All)
sysctl hw.model
hw.model: Intel(R) Xeon(R) CPU E5335 @ 2.00GHzhttp://ark.intel.com/products/28443/Inte...33-MHz-FSB
Installation
cobalthost is using their own Panel, which fulfills the basic needs, shows cpu/ram/hdd usage and offers the usual suspects like reboot/reinstall etc.
CPU Usage is always wrong in the panel, mobil version does not work. Installation works, which is most important.
One can choose to install on DC-2 or DC-3, since nowhere i can see the utilization of either node or see what hardware is in it it's is a gamble.
I first chose DC-3, which was a mistake. It is so slow that there can be only 2 reasons for it. a) it has like hundreds of VMs running, or b) the hw is dying (as in faulty), possible the hdds.
Booting a VM takes up to 5 minutes, installing ~30 minutes.
So i reinstalled on DC-2, which works better, boot and install times are normal. Installation went without issues. Their install script leaves a /bin/npasswd script back and a non working rc.local entry in case of debian. On FreeBSD i could not find any artifacts but 2 installed packages.
There is a slight installation bug with FreeBSD, which i already reported. (swap missing), also it installs on UFS2 instead of ZFS, which is a bummer but should do for most tasks.
One cool with with cobalthost is that you can have as many VPSs as you have resources for, so with 2 cores you can do 2 VPSs. Very cool.
Benchmark
Since i am running FreeBSD the usually FreeVPS benchmark script won't work, so i had to improvise and used unixbench.
Unixbench:
Code: (Select All)
BYTE UNIX Benchmarks (Version 4.1.0)
System -- cobalt
Start Benchmark Run: Sat Mar 5 14:18:55 CST 2016
1 interactive users.
2:18PM up 18:48, 1 user, load averages: 0.32, 0.24, 0.17
-r-xr-xr-x 1 root wheel 142208 Mar 4 19:21 /bin/sh
/bin/sh: ELF 64-bit LSB executable, x86-64, version 1 (FreeBSD), dynamically linked, interpreter /libexec/ld-elf.so.1, for FreeBSD 10.2, stripped
/dev/da0p2 49589108 3605016 42016964 8% /
Dhrystone 2 using register variables 10248240.4 lps (10.0 secs, 10 samples)
Double-Precision Whetstone 1581.4 MWIPS (11.5 secs, 10 samples)
System Call Overhead 352145.8 lps (10.0 secs, 10 samples)
Pipe Throughput 90979.2 lps (10.0 secs, 10 samples)
Pipe-based Context Switching 5820.0 lps (10.0 secs, 10 samples)
Process Creation 293.4 lps (30.0 secs, 3 samples)
ecl Throughput 127.2 lps (29.8 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 337529.0 KBps (30.0 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 21709.0 KBps (30.0 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 20583.0 KBps (30.0 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 107441.0 KBps (30.0 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 24680.0 KBps (30.0 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks 19831.0 KBps (30.0 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks 731668.0 KBps (30.0 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks 20279.0 KBps (30.0 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks 20593.0 KBps (30.0 secs, 3 samples)
Shell Scripts (1 concurrent) 341.4 lpm (60.1 secs, 3 samples)
Shell Scripts (8 concurrent) 55.8 lpm (60.1 secs, 3 samples)
Shell Scripts (16 concurrent) 25.9 lpm (60.1 secs, 3 samples)
Arithmetic Test (type = short) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = int) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = long) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = float) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = double) 1.0 lps (0.0 secs, 3 samples)
Arithoh 1.0 lps (0.0 secs, 3 samples)
C Compiler Throughput 12610.5 lpm (60.0 secs, 3 samples)
Dc: sqrt(2) to 99 decimal places 12301.0 lpm (30.1 secs, 3 samples)
Recursion Test--Tower of Hanoi 111125.6 lps (20.0 secs, 3 samples)
INDEX VALUES
TEST BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 10248240.4 878.2
Double-Precision Whetstone 55.0 1581.4 287.5
Execl Throughput 43.0 127.2 29.6
File Copy 1024 bufsize 2000 maxblocks 3960.0 20583.0 52.0
File Copy 256 bufsize 500 maxblocks 1655.0 19831.0 119.8
File Copy 4096 bufsize 8000 maxblocks 5800.0 20593.0 35.5
Pipe Throughput 12440.0 90979.2 73.1
Pipe-based Context Switching 4000.0 5820.0 14.6
Process Creation 126.0 293.4 23.3
Shell Scripts (8 concurrent) 6.0 55.8 93.0
System Call Overhead 15000.0 352145.8 234.8
=========
FINAL SCORE 80.3Now that looks kind of horrible. As a small comparison a FreeBSD VM on an older core i3 under load in production had a Final Score of "231". FreeBSD VM in an older Xeon also in production with load on it did "362".
ioping:
Code: (Select All)
ioping .
4 KiB from . (ufs /dev/da0p2): request=1 time=11.1 ms
4 KiB from . (ufs /dev/da0p2): request=2 time=2.89 ms
4 KiB from . (ufs /dev/da0p2): request=3 time=2.71 ms
4 KiB from . (ufs /dev/da0p2): request=4 time=142.8 ms
4 KiB from . (ufs /dev/da0p2): request=5 time=268.2 ms
4 KiB from . (ufs /dev/da0p2): request=6 time=13.5 ms
4 KiB from . (ufs /dev/da0p2): request=7 time=388.9 ms
4 KiB from . (ufs /dev/da0p2): request=8 time=2.34 ms
4 KiB from . (ufs /dev/da0p2): request=9 time=47.9 ms
4 KiB from . (ufs /dev/da0p2): request=10 time=4.28 ms
4 KiB from . (ufs /dev/da0p2): request=11 time=7.67 ms
4 KiB from . (ufs /dev/da0p2): request=12 time=17.1 ms
4 KiB from . (ufs /dev/da0p2): request=13 time=21.1 ms
4 KiB from . (ufs /dev/da0p2): request=14 time=2.03 ms
4 KiB from . (ufs /dev/da0p2): request=15 time=2.55 ms
4 KiB from . (ufs /dev/da0p2): request=16 time=132.4 msAs you can see, it is worse than on a raspberry pi on sdcards, the 300ms+ peaks are especially bad.
dd:
Code: (Select All)
dd if=/dev/zero of=test bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes transferred in 232.203143 secs (4624149 bytes/sec)Now that is also right up the raspberry pis disks speed as well. In all fairness, doing the test during different hour yield better results at around 12-30MB/s
Summary: Disk speeds and latency are bad.
CPU:
I tested CPU speed with openssl speed which every sysem has for comparison:
Code: (Select All)
openssl speed
Doing mdc2 for 3s on 16 size blocks: 1163607 mdc2's in 3.04s
Doing mdc2 for 3s on 64 size blocks: 324635 mdc2's in 3.01s
Doing mdc2 for 3s on 256 size blocks: 81461 mdc2's in 3.01s
Doing mdc2 for 3s on 1024 size blocks: 18306 mdc2's in 2.99s
Doing mdc2 for 3s on 8192 size blocks: 1650 mdc2's in 3.03s
Doing md4 for 3s on 16 size blocks: 3271393 md4's in 2.99s
Doing md4 for 3s on 64 size blocks: 2790786 md4's in 3.00s
Doing md4 for 3s on 256 size blocks: 1562874 md4's in 3.01s
Doing md4 for 3s on 1024 size blocks: 904403 md4's in 3.01s
Doing md4 for 3s on 8192 size blocks: 120088 md4's in 3.01s
Doing md5 for 3s on 16 size blocks: 1729218 md5's in 3.00s
Doing md5 for 3s on 64 size blocks: 1923173 md5's in 3.02s
Doing md5 for 3s on 256 size blocks: 1734881 md5's in 3.02s
Doing md5 for 3s on 1024 size blocks: 658213 md5's in 3.02s
Doing md5 for 3s on 8192 size blocks: 105286 md5's in 3.01s
Doing hmac(md5) for 3s on 16 size blocks: 2335442 hmac(md5)'s in 3.02s
Doing hmac(md5) for 3s on 64 size blocks: 1940267 hmac(md5)'s in 3.05s
Doing hmac(md5) for 3s on 256 size blocks: 1241445 hmac(md5)'s in 3.00s
Doing hmac(md5) for 3s on 1024 size blocks: 399851 hmac(md5)'s in 3.00s
Doing hmac(md5) for 3s on 8192 size blocks: 76959 hmac(md5)'s in 3.03s
Doing sha1 for 3s on 16 size blocks: 2647621 sha1's in 3.00s
Doing sha1 for 3s on 64 size blocks: 2665564 sha1's in 3.00s
Doing sha1 for 3s on 256 size blocks: 1675455 sha1's in 3.02s
Doing sha1 for 3s on 1024 size blocks: 538815 sha1's in 3.01s
Doing sha1 for 3s on 8192 size blocks: 86528 sha1's in 3.02s
Doing sha256 for 3s on 16 size blocks: 3355346 sha256's in 3.00s
Doing sha256 for 3s on 64 size blocks: 1753528 sha256's in 3.00s
Doing sha256 for 3s on 256 size blocks: 776093 sha256's in 3.00s
Doing sha256 for 3s on 1024 size blocks: 226618 sha256's in 3.02s
Doing sha256 for 3s on 8192 size blocks: 38635 sha256's in 3.02s
Doing sha512 for 3s on 16 size blocks: 2680159 sha512's in 2.97s
Doing sha512 for 3s on 64 size blocks: 2317665 sha512's in 3.00s
Doing sha512 for 3s on 256 size blocks: 1088552 sha512's in 2.98s
Doing sha512 for 3s on 1024 size blocks: 361355 sha512's in 3.02s
Doing sha512 for 3s on 8192 size blocks: 44872 sha512's in 3.02sNetwork:
Now the specs say 100mbit, but in reality it is connected with gbit, which is nice. Speeds inside Canada are good, ok to the USA and not so good to Europe but that was to be expected.
Code: (Select All)
Selecting best server based on latency...
Hosted by Rack and Data (Kitchener, ON) [4.61 km]: 50.804 ms
Testing download speed........................................
Download: 261.26 Mbit/s
Testing upload speed..................................................
Upload: 66.44 Mbit/sCode: (Select All)
Hosted by Optimum Online (New York City, NY) [614.16 km]: 78.655 ms
Testing download speed........................................
Download: 62.83 Mbit/s
Testing upload speed..................................................
Upload: 46.89 Mbit/sCode: (Select All)
Hosted by Vodafone DE (Frankfurt) [6418.82 km]: 250.362 ms
Testing download speed........................................
Download: 20.78 Mbit/s
Testing upload speed..................................................
Upload: 15.82 Mbit/sMy VPN does 15mbit (compressed) to my home link, which is enough for my VPN related activities. Also enough for mumble/teamspeak or a small private website.
VPS Usage
I mainly use it for playing around with Operatingsystems and run a mumble and an OpenVPN server, both private for myself. Maybe i'll use to backup some stuff, since the 50GB disk space is nice.
Reliability
Can not say much about that yet, for now it has been stable, if it ever goes down i will update this thread.
Conclussion
Well that is a mixed bag, it is very old, dying hardware and probably over sold into oblivion. So if you are looking for a VPS where you actually have to run something that needs power or disk speeds from this century you are out of luck. A quad core CPU for half their customers(DC-3 has 6 cores i think) is simply not enough, also the apparent lack of a storage for the VMs other than the included HDDs on the old server shows, very much so. If you are looking for a VPS to do some private stuff, test your code, learn linux, run some small daemons or anything similar like that it does its job, after all it's free and for free it's good, can't really complain. For anyone looking to invest their own money i would suggest to look for another hoster.
greetings
p.s. redid all test with debian, same result, does not seem to be OS related.