Community Review: KiloServe 365 MB KVM (Yearly)
KiloServe is a company I've wanted to review for a while as they seem to have a fairly good reputation and have been in business for a while. Since I had some free time I decided to bite the bullet and purchase a KVM365 plan for testing and hosting some websites.
Disclaimer: I am in no way affiliated with KiloServe. This review was requested by staff from KiloServe but the service was fully paid for by myself.
Basics:
The VPS plan is based in LA, USA and comes with 365 MB of RAM, 365 GB of bandwidth (monthly) and 15 GB of disk space. KVM plans are managed by a standard SolusVM install (with SSL) and include commonly used templates such as CentOS (5 and 6), Ubuntu and Debian (unfortunately Debian 6 was not included, although this was soon amended with a ticket). On the downside both torrenting and IRC are prohibited making the servers a little bit less useful to some users.
Support:
Initial setup was very fast and the server was running in under five minutes from payment however the welcome email did contain plain text passwords which could be considered a security risk by some (personally I just change the passwords if this is the case). Since the KVM ISO list did not contain Debian 6 I put in a ticket requesting it. Within 10 minutes I had a reply stating that it was being added and soon after that it appeared on my ISO list.
Setup:
After the ISO was added I proceded to install Debian 6 as normal (installing nothing but the SSH server). After it was installed I used my minimisation script to clean out any unneeded packages and to set up SSH login protection. One thing I did notice was that commands sent from SolusVM were processed rather slowly (my KVM often took about a minute to power off after clicking the button).
Defaults:
Default RAM usage was pretty low for a KVM server with about 13 MB used on a stock Debian 6 install:
[email protected]:~# free -m total used free shared buffers cached Mem: 356 52 304 0 26 12 -/+ buffers/cache: 13 343 Swap: 2007 0 2007
Default disk usage was also fairly low:
[email protected]:~# df -h Filesystem Size Used Avail Use% Mounted on /dev/sda1 1.9G 358M 1.5G 20% / tmpfs 179M 0 179M 0% /lib/init/rw udev 174M 112K 174M 1% /dev tmpfs 179M 0 179M 0% /dev/shm /dev/sda2 12G 156M 11G 2% /home
Basic Information:
/proc/cpuinfo showed 4 cores which matched the plan description:
[email protected]:~# cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : QEMU Virtual CPU version (cpu64-rhel6) stepping : 3 cpu MHz : 3058.352 cache size : 4096 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 4 wp : yes flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm pni cx16 hypervisor lahf_lm bogomips : 6116.70 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : QEMU Virtual CPU version (cpu64-rhel6) stepping : 3 cpu MHz : 3058.352 cache size : 4096 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 4 wp : yes flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm pni cx16 hypervisor lahf_lm bogomips : 6116.70 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: processor : 2 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : QEMU Virtual CPU version (cpu64-rhel6) stepping : 3 cpu MHz : 3058.352 cache size : 4096 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 4 wp : yes flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm pni cx16 hypervisor lahf_lm bogomips : 6116.70 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: processor : 3 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : QEMU Virtual CPU version (cpu64-rhel6) stepping : 3 cpu MHz : 3058.352 cache size : 4096 KB fdiv_bug : no hlt_bug : no f00f_bug : no coma_bug : no fpu : yes fpu_exception : yes cpuid level : 4 wp : yes flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm pni cx16 hypervisor lahf_lm bogomips : 6116.70 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management:
/proc/meminfo showed some standard results:
[email protected]:~# cat /proc/meminfo MemTotal: 365168 kB MemFree: 311516 kB Buffers: 27280 kB Cached: 12616 kB SwapCached: 0 kB Active: 31084 kB Inactive: 11516 kB Active(anon): 2708 kB Inactive(anon): 104 kB Active(file): 28376 kB Inactive(file): 11412 kB Unevictable: 0 kB Mlocked: 0 kB HighTotal: 0 kB HighFree: 0 kB LowTotal: 365168 kB LowFree: 311516 kB SwapTotal: 2056184 kB SwapFree: 2056184 kB Dirty: 0 kB Writeback: 0 kB AnonPages: 2708 kB Mapped: 3456 kB Shmem: 116 kB Slab: 6468 kB SReclaimable: 2560 kB SUnreclaim: 3908 kB KernelStack: 704 kB PageTables: 188 kB NFS_Unstable: 0 kB Bounce: 0 kB WritebackTmp: 0 kB CommitLimit: 2238768 kB Committed_AS: 8144 kB VmallocTotal: 654348 kB VmallocUsed: 6572 kB VmallocChunk: 639664 kB HardwareCorrupted: 0 kB HugePages_Total: 0 HugePages_Free: 0 HugePages_Rsvd: 0 HugePages_Surp: 0 Hugepagesize: 4096 kB DirectMap4k: 9196 kB DirectMap4M: 364544 kB
Inode allocation was decent (this was however controlled by my install so KiloServe has nothing to do with this):
[email protected]:~# df -i Filesystem Inodes IUsed IFree IUse% Mounted on /dev/sda1 122160 15279 106881 13% / tmpfs 45646 3 45643 1% /lib/init/rw udev 44540 516 44024 2% /dev tmpfs 45646 1 45645 1% /dev/shm /dev/sda2 732960 15 732945 1% /home
vmstat showed that the system was under little load:
[email protected]:~# vmstat procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu---- r b swpd free buff cache si so bi bo in cs us sy id wa 0 0 0 311516 27280 12616 0 0 27 3 17 10 0 0 100 0
Comments
Tests:
Each test was run three times and the middle ranked test was picked.
The Cachefly speed test showed great results:
Ping Tests (as with DMB Hosting, IPv6 did not appear to be included):
Disk IO on KiloServe is unreal, the fastest I have ever seen (on a VPS):
IOPing showed good results as well:
Geekbench results were very good:
Finally, Unixbench also showed great results:
======================================================================== BYTE UNIX Benchmarks (Version 5.1.3) System: kilo: GNU/Linux OS: GNU/Linux -- 2.6.32-5-686 -- #1 SMP Mon Oct 3 04:15:24 UTC 2011 Machine: i686 (unknown) Language: en_US.utf8 (charmap="ANSI_X3.4-1968", collate="ANSI_X3.4-1968") CPU 0: QEMU Virtual CPU version (cpu64-rhel6) (6116.7 bogomips) x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET CPU 1: QEMU Virtual CPU version (cpu64-rhel6) (6116.7 bogomips) x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET CPU 2: QEMU Virtual CPU version (cpu64-rhel6) (6116.7 bogomips) x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET CPU 3: QEMU Virtual CPU version (cpu64-rhel6) (6116.7 bogomips) x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET 19:54:06 up 22 min, 1 user, load average: 0.00, 0.05, 0.03; runlevel 2 ------------------------------------------------------------------------ Benchmark Run: Fri Nov 04 2011 19:54:06 - 20:22:25 4 CPUs in system; running 1 parallel copy of tests Dhrystone 2 using register variables 16604288.8 lps (10.0 s, 7 samples) Double-Precision Whetstone 3236.0 MWIPS (10.1 s, 7 samples) Execl Throughput 4074.4 lps (29.8 s, 2 samples) File Copy 1024 bufsize 2000 maxblocks 822195.8 KBps (30.0 s, 2 samples) File Copy 256 bufsize 500 maxblocks 215533.1 KBps (30.0 s, 2 samples) File Copy 4096 bufsize 8000 maxblocks 1747124.6 KBps (30.0 s, 2 samples) Pipe Throughput 1397073.3 lps (10.0 s, 7 samples) Pipe-based Context Switching 341646.2 lps (10.0 s, 7 samples) Process Creation 8919.0 lps (30.0 s, 2 samples) Shell Scripts (1 concurrent) 7366.8 lpm (60.0 s, 2 samples) Shell Scripts (8 concurrent) 2957.4 lpm (60.0 s, 2 samples) System Call Overhead 1317068.5 lps (10.0 s, 7 samples) System Benchmarks Index Values BASELINE RESULT INDEX Dhrystone 2 using register variables 116700.0 16604288.8 1422.8 Double-Precision Whetstone 55.0 3236.0 588.4 Execl Throughput 43.0 4074.4 947.5 File Copy 1024 bufsize 2000 maxblocks 3960.0 822195.8 2076.3 File Copy 256 bufsize 500 maxblocks 1655.0 215533.1 1302.3 File Copy 4096 bufsize 8000 maxblocks 5800.0 1747124.6 3012.3 Pipe Throughput 12440.0 1397073.3 1123.0 Pipe-based Context Switching 4000.0 341646.2 854.1 Process Creation 126.0 8919.0 707.9 Shell Scripts (1 concurrent) 42.4 7366.8 1737.5 Shell Scripts (8 concurrent) 6.0 2957.4 4929.1 System Call Overhead 15000.0 1317068.5 878.0 ======== System Benchmarks Index Score 1338.2 ------------------------------------------------------------------------ Benchmark Run: Fri Nov 04 2011 20:22:25 - 20:50:45 4 CPUs in system; running 4 parallel copies of tests Dhrystone 2 using register variables 66086854.8 lps (10.0 s, 7 samples) Double-Precision Whetstone 12921.1 MWIPS (10.1 s, 7 samples) Execl Throughput 23541.8 lps (29.9 s, 2 samples) File Copy 1024 bufsize 2000 maxblocks 764702.5 KBps (30.0 s, 2 samples) File Copy 256 bufsize 500 maxblocks 171526.5 KBps (30.0 s, 2 samples) File Copy 4096 bufsize 8000 maxblocks 2390502.8 KBps (30.0 s, 2 samples) Pipe Throughput 5511619.3 lps (10.0 s, 7 samples) Pipe-based Context Switching 1269416.2 lps (10.0 s, 7 samples) Process Creation 76388.6 lps (30.0 s, 2 samples) Shell Scripts (1 concurrent) 27024.5 lpm (60.0 s, 2 samples) Shell Scripts (8 concurrent) 3557.9 lpm (60.0 s, 2 samples) System Call Overhead 4785736.0 lps (10.0 s, 7 samples) System Benchmarks Index Values BASELINE RESULT INDEX Dhrystone 2 using register variables 116700.0 66086854.8 5663.0 Double-Precision Whetstone 55.0 12921.1 2349.3 Execl Throughput 43.0 23541.8 5474.8 File Copy 1024 bufsize 2000 maxblocks 3960.0 764702.5 1931.1 File Copy 256 bufsize 500 maxblocks 1655.0 171526.5 1036.4 File Copy 4096 bufsize 8000 maxblocks 5800.0 2390502.8 4121.6 Pipe Throughput 12440.0 5511619.3 4430.6 Pipe-based Context Switching 4000.0 1269416.2 3173.5 Process Creation 126.0 76388.6 6062.6 Shell Scripts (1 concurrent) 42.4 27024.5 6373.7 Shell Scripts (8 concurrent) 6.0 3557.9 5929.9 System Call Overhead 15000.0 4785736.0 3190.5 ======== System Benchmarks Index Score 3680.4Conclusion:
In conclusion I believe that KiloServe is a very good host with great performance. Ahe prices and response times for support are also very competitive and I would not hesitate to recommend KiloServe to anyone in need of a quality VPS. Finally, their TOS protects their service from abusers creating a more stable and speedy enviroment for legitimate users.
Please give KiloServe a try and report on findings! Thanks for reading, tips and suggestions are appreciated!
I get a bit more on Kazila VPS :P
I wonder if it's possible to get 128 MB KVM VPS for $3 monthly. Thank you for review.
I should do a review of them as well, I hear that they are good!
What is the cost of this VPS used for testing?
$39.49 annually.
+1 for the review. Nicely formatted.
I've been using them for 2 months now... making it my main VPS after the eNetSouth event. I've been very satisfied so far.
sigh was excited to sign up until I realized there was a $3 setup fee with no refund when it's a yearly term![:\](https://lowendtalk.com/resources/emoji/confused.png ":\")
I don't understand it. Usually setup fees are a good way to keep some cash when ppl sign up and cancel a few days later. Paying yearly with no way out is downright scary.
Nice
Any idea what processors are using? Must be at least some Nehalems?
If you view the benchmarks (337 MB/s) and the price ($36.50/year), this is both one of the fastest and lowest priced KVM VPS on the market.... even when you count in the $2.99 setup fee.
Unfortunately we will not be getting into the 128MB market. 365MB RAM is the lowest we will go.
We've been in business for almost 5 years and have provided refunds that our TOS says aren't available because we know when a "refund" is the right thing to do.... regardless of the rules.
Westmere-EP
SAS2/SATA3 ? Very impressive!
Thanks Tim.
Good guess, they are indeed SAS2 (6Gb/s) drives....thankfully, we completed these nodes before the hard drive shortage.
BTW, I had a client just this morning tell me that they love your network connectivity/routes.
Well it's $3.04 x 12 = 36.48 / annually. Plus a $2.99 setup fee. I wouldn't guess the setup fee is charged annually
I have absolutely no problem paying a one-time setup fee to a reputable provider. Kiloserv is a reputable provider. And what's $3? Maybe 5 minutes of paid time for an experienced sysadmin? If a setup fee helps keep the business viable, go for it!
The last time I trusted a "reputable" provider (starts with an "e") and they disappeared after 3 days of payment. It's not so much about the money (2 of my vps are not leb range at all so I'm not afraid to pay for quality) but it's more about this promise of a full year service that makes it a risky proposition.
Now @kilsoserve, I'm not trying to lump you with other providers so please don't get defensive here. I can tell you are a reputable business but I guess I'm tired of being unable to test a provider without throwing down a good chunk of money. It's like "You don't trust your customers to commit so why should I trust you?". Though I'm quite sure someone would say "why don't you sign up for the regular plans?" sigh
Anyway, I may or may not sign up some time later (if the LEB plan is still offered). Right now, I have enough LEBs to play with and enough "secondary" dns servers and backups of backups so I'll give it a rest :P
Hi tortau,
Thanks for being so polite about it all.
Truth is, we didn't have any of the "cool" yearly plans everybody else has so we made this one up.
This is our only plan based on a year, all of our other plans, are monthly.
Taking a look at their site, they have a nice description of their nodes:
http://kiloserve.com/48-core-nodes
Pretty impressive stuff. They have room to expand on the memory front too if they move to 8GB sticks of ram.
That's awesome! Lol at the Windows task manager in that picture...Very impressive. One downside I guess is if there's hardware failure. I wouldn't want to be the guy trying to figure out which of the 8 processors died, or which of the 32 ram sticks is bad.
@maxexcloo well done. thanks for posting your benchmark data.
I too have one of their 365's. Very impressive for LEB price imho. Great work @kiloserve
@kiloserve Heh, thanks for the reply. I was kinda worrying about coming home to a long thread of angry comments so I feel somewhat cheated now. Haha :P
Anyway, I did notice the monthly plans but I do prefer KVM (I'm hooked on them actually - don't ask why) and it was the only option on your site
One of the most detailed reviews I've seen, must be refering to it whenever I come across them.
More reviews of that type http://www.96mb.com/
Thanks! If you're interested have a look at some of my other reviews here: http://www.lowendtalk.com/profile/discussions/38/maxexcloo
After reading his comments on this thread, I bought kvm365 from him. I'm very satisfied with that box, long uptime, very good disk and network I/O, and haven't found any troubles yet.
BTW, sad story, I've just paid 2nd month for eNetSouth 2GB plan, then my disk suddenly full 100% and can't access they SolusVM page T_T
another guess.. What kind disks are you using? 15k rpm? And what raid setup, with how many disks? last time i only can reach ~ 200-250 MB/s with the "same" hard drive..
Hi Monst3r
It's a real pain getting speed out of KVM. To be perfectly honest, our old KVM/Xen-HVM offerings were embarrassingly slow and it took us many tries before we were able to achieve decent speeds.
With OpenVZ and Xen, you can use 10k drives and still get 300 MB/s+ write speeds but you need 15K drives to come close to that on KVM. The extra RPM's overcome the I/O latency inherent to KVM due to the full virt.
We now use 15K SAS 6Gb/s drives for KVM, there's really no other way to get good consistent speed out of KVM with only minor degradation as more VPS are added on.
BTW, I tried to get one of your NL KVM VPS just a day or two ago but you were out of stock. Please PM me if you get stock in and if you remember; I'll check back every so often too
That's very speedy
Is that XenPV or KVM?
OpenVZ's and XenPV's will have higher disk I/O speeds than KVM or Xen-HVM due to native disk access as opposed to full virtualization. I think most KVM/XenHVM providers will agree with me that achieving and maintaining 300MB/s+ write speeds on KVM/XenHVM is very difficult.
It's XenPV.
I put the server to use by hosting some websites and an EP my friends released