Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


The Real NVMe Deal 80GB NVMe and 6GB ram for 7$/month - servaRICA
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

The Real NVMe Deal 80GB NVMe and 6GB ram for 7$/month - servaRICA

servarica_haniservarica_hani Member, Patron Provider

Hello Everyone,

Today we have our first NVMe based offer.

And since it is the first offer we decided to make it as beefy as possible .

with 6GB RAM , 4 cores and 80GB NVMe disk you will enough resources to run your application with no issues.

This is Xen based VPS (similar to KVM if you are unfamiliar with Xen)

The Offers:

NVME Cheetah

NVME Cheetah 
4x CPU cores 
6GB RAM 
80GB NVMe disk
unlimited transfer
100mbps vps port 
1x IPv4
IPv6 available by request
$7$/month

Order here https://servarica.com/clients/cart.php?a=add&pid=408

Offer Features

This offer comes with

  • CPU is E5-2650
  • Xenserver based
  • 7 days refund policy with no question asked
  • 99.9% SLA
  • We offer true 24/7 livechat support
  • We are a stable host that has been in business since 2010 with good reviews
  • We own all our hardware and we manage our own network with our own AS number. All these offers are based on our greater Montreal Datacenter

SLA

99.9%: normal operation
99.0% or more: 15% refund
97.0% or more: 30% refund
95.0% or more: 50% refund
Less than 95%: 100% refund

Network information

Montreal, Canada
Test IPv4: 162.250.190.17
Test file: http://162.250.190.17/file_100MB.bin

search about us if you are not familiar with us we have been around for a long time.

Thanks
Hani
servaRICA.com

Thanked by 2alilet okgoogle
«13

Comments

  • Windows ( lisenced/trial )possible on this offer?

  • any benchmark for this?

  • Nice offer

  • Do you have free internal directadmin licenses?

  • A few benchmarks would be helpful. vCPU and RAM specs alone don't tell much about the host computing power and network bandwidth.

    cryptsetup -c aes-xts-plain64 -s 256 benchmark
    nench.sh
    sysbench cpu --cpu-max-prime=20000 --threads=4 --time=60 run

    Thanked by 1plumberg
  • servarica_haniservarica_hani Member, Patron Provider

    @jbuggie said:
    A few benchmarks would be helpful. vCPU and RAM specs alone don't tell much about the host computing power and network bandwidth.

    here you go

    sysbench cpu --cpu-max-prime=20000 --threads=4 --time=60 run

    sysbench cpu --cpu-max-prime=20000 --threads=4 --time=60 run
    sysbench 1.0.11 (using system LuaJIT 2.1.0-beta3)
    
    Running the test with following options:
    Number of threads: 4
    Initializing random number generator from current time
    
    
    Prime numbers limit: 20000
    
    Initializing worker threads...
    
    Threads started!
    
    .CPU speed:
        events per second:  1472.38
    
    General statistics:
        total time:                          60.0016s
        total number of events:              88348
    
    Latency (ms):
            min:                                  2.61
            avg:                                  2.72
            max:                                 16.39
            95th percentile:                      2.91
            sum:                             239891.83
    
    Threads fairness:
        events (avg/stddev):           22087.0000/14.85
        execution time (avg/stddev):   59.9730/0.00
    

    cryptsetup -c aes-xts-plain64 -s 256 benchmark

    cryptsetup -c aes-xts-plain64 -s 256 benchmark
    # Tests are approximate using memory only (no storage IO).
    #     Algorithm | Key |  Encryption |  Decryption
            aes-xts   256b   833.2 MiB/s   926.0 MiB/s
    

    nench.sh

    -------------------------------------------------
    nench.sh v2019.07.20 -- https://git.io/nench.sh
    benchmark timestamp:    2019-09-03 04:35:32 UTC
    -------------------------------------------------
    
    Processor:    Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz
    CPU cores:    4
    Frequency:    2593.766 MHz
    RAM:          5.7G
    Swap:         1.0G
    Kernel:       Linux 3.10.0-123.13.2.el7.x86_64 x86_64
    
    Disks:
    xvda     80G  SSD
    
    CPU: SHA256-hashing 500 MB
        2.718 seconds
    CPU: bzip2-compressing 500 MB
        8.008 seconds
    CPU: AES-encrypting 500 MB
        3.021 seconds
    
    
    ioping: seek rate
        min/avg/max/mdev = 118.8 us / 204.3 us / 3.09 ms / 61.5 us
    ioping: sequential read speed
        generated 7.31 k requests in 5.00 s, 1.79 GiB, 1.46 k iops, 365.6 MiB/s
    
    dd: sequential write speed
        1st run:    298.50 MiB/s
        2nd run:    290.87 MiB/s
        3rd run:    280.38 MiB/s
        average:    289.92 MiB/s
    
    
    IPv4 speedtests
        your IPv4:    104.152.209.xxxx
    
        Cachefly CDN:         11.56 MiB/s
        Leaseweb (NL):        10.24 MiB/s
        Softlayer DAL (US):   9.77 MiB/s
        Online.net (FR):      9.92 MiB/s
        OVH BHS (CA):         11.46 MiB/s
    
    No IPv6 connectivity detected
    

    Notes:
    1- I did run the tests on the most crowded node so far intentionally so you can get real feeling of the performance
    2- the network is limited to 100mbps on the vm thats why the max download is always around 10MB/s

    Thanked by 1uptime
  • You sure that’s NVMe?

  • servarica_haniservarica_hani Member, Patron Provider

    @corbpie said:
    You sure that’s NVMe?

    yes and it is very fast as well
    here are the specs if you are interested
    https://www.samsung.com/semiconductor/global.semi.static/PM1725_NVMe_PCIe_SSD_Samsung_2016-1.pdf

    the issue is in xenserver in general, their io path is not optimized for single vm performance .
    I mean running 50 vms on the node we can saturate the disk easily but a single vm is always limited .

    So you can think of the numbers you see here as guaranteed numbers (it will never go less than that)

    by end of this year citrix will release new storage drivers based on their new code base SMAPIv3 that is supposed to fix this issue among others and will allow single vm to perform much better than this in terms of disk io

    on the other hand the reliability and the mature of the xen echo system is worth it for us to have io performance hit

    Thanked by 1coreflux
  • rchurchrchurch Member
    edited September 2019

    Is it Xen HVM or Xen PV?

    It is long since I heard this terms mentioned :)

    Thanked by 1Falzo
  • That's a decent performance for 4-core offering at this price point. I've seen a lot worse. Anyone using dmcrypt on their VPS is going to be further limited by aes encryption performance. Any disk operation will quickly become CPU-bound at a much lower rate (dmcrypt is a bit faster than nench.sh 500MB/3s+ but not much). NVMe or regular SSD would not make much of a difference.

    Any plan to go for limited bw at higher port speed? How about traffic to your other storage VM?

  • Nvme vps should get a 2x portspeed bump at least .

    Thanked by 1cybertech
  • servarica_haniservarica_hani Member, Patron Provider
    edited September 2019

    @rchurch said:
    Is it Xen HVM or Xen PV?

    It is long since I heard this terms mentioned :)

    it is XEN PVHVM or what they call it PV over HVM

    basically you get the good from both worlds

    it is an HVM VM with PV drivers to increase network speed specifically

    soon citrix will finally release PVH on xenserver at that point we will be the kings of virtualization world :)

  • @Hani said:

    @rchurch said:
    Is it Xen HVM or Xen PV?

    It is long since I heard this terms mentioned :)

    it is XEN PVHVM or what they call it PV over HVM

    basically you get the good from both worlds

    it is an HVM VM with PV drivers to increase network speed specifically

    soon citrix will finally release PVH on xenserver at that point we will be the kings of virtualization world :)

    Can you upload your own ISOs?

  • servarica_haniservarica_hani Member, Patron Provider

    @rchurch said:
    Can you upload your own ISOs?

    yes just open a ticket with your iso url and we will mount it for you then you can install it yourself

  • dd: sequential write speed
    1st run: 298.50 MiB/s
    2nd run: 290.87 MiB/s
    3rd run: 280.38 MiB/s
    average: 289.92 MiB/s

    Limiting iops?

  • ManishPantManishPant Member, Host Rep

    Any chance of you adding windows original license to it also in future

  • servarica_haniservarica_hani Member, Patron Provider

    @jordynegen11 said:

    dd: sequential write speed
    1st run: 298.50 MiB/s
    2nd run: 290.87 MiB/s
    3rd run: 280.38 MiB/s
    average: 289.92 MiB/s

    Limiting iops?

    no as i explained it is xenserver thing we didnt limit anything
    also dd test is not giving you real life performance the vm is performing really well

  • servarica_haniservarica_hani Member, Patron Provider

    @ManishPant said:
    Any chance of you adding windows original license to it also in future

    I dont think so because we get windows license per physical server and due to high ram on these plans the number of vps per server is low which make the license cost high on the end user

  • jsgjsg Member, Resident Benchmarker

    @newbiemasih said:
    any benchmark for this?

    I'd be willing to do a benchmark and review if @Hani is interested in it.

  • servarica_haniservarica_hani Member, Patron Provider

    @jsg said:

    @newbiemasih said:
    any benchmark for this?

    I'd be willing to do a benchmark and review if @Hani is interested in it.

    go ahead make an order and dont pay
    then update me by message of your order and i will accept it

    you will have 1 day to do benchmarks

    Thanked by 2ITLabs kkrajk
  • ManishPantManishPant Member, Host Rep

    @Hani said:

    @jsg said:

    @newbiemasih said:
    any benchmark for this?

    I'd be willing to do a benchmark and review if @Hani is interested in it.

    go ahead make an order and dont pay
    then update me by message of your order and i will accept it

    you will have 1 day to do benchmarks

    That's great @jsg let us know the review

  • jsgjsg Member, Resident Benchmarker

    @Hani said:
    go ahead make an order and dont pay
    then update me by message of your order and i will accept it

    you will have 1 day to do benchmarks

    OK, thanks. I'll contact you.

  • I propose you should provide a test server to @Falzo ... If his evaluations find it good, you will have your name and offer taken more seriously.

  • servarica_haniservarica_hani Member, Patron Provider

    @mehargags said:
    I propose you should provide a test server to @Falzo ... If his evaluations find it good, you will have your name and offer taken more seriously.

    I dont mind
    if @Falzo want to test the vps i can give him free vps for 3 days to do his testing

  • Nice offer! any limitations on core usage ?

  • servarica_haniservarica_hani Member, Patron Provider

    @Kit said:
    Nice offer! any limitations on core usage ?

    you can use up to 25% of the 4 cores (= 1 full core 100%) all the time
    you can use all the cores 100% for 2 hours burst with no issues
    if there is need of the cpus after 2 hours you will be limited to 25% for 2 hours and then the limiting will be lifted (automated)

  • Do you offer DirectAdmin license? Price?

  • servarica_haniservarica_hani Member, Patron Provider

    @LowVM said:
    Do you offer DirectAdmin license? Price?

    yes it cost 6$/month
    we have added it as an option while ordering few hours ago since other users asked for it as well

  • Is it possible to move my ssd vps(same price) to nvme offer + extra directadmin license?

  • jsgjsg Member, Resident Benchmarker
    edited September 2019

    OK, here's the test/benchmarks results:

    Machine: amd64, Arch.: amd64, Model: Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz
    OS, version: FreeBSD 12.0, Mem.: 5.977 GB
    CPU - Cores: 4, Family/Model/Stepping: 6/62/4
    Cache: 32K/32K L1d/L1i, 256K L2, 20M L3
    Std. Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
              pse36 cflsh acpi mmx fxsr sse sse2 ss htt sse3 pclmulqdq ssse3 cx16
              pcid sse4_1 sse4_2 x2apic popcnt tsc_deadline aes xsave osxsave avx
              f16c rdrnd hypervisor
    Ext. Flags: fsgsbase smep erms syscall nx pdpe1gb rdtscp lm lahf_lm
    
    --- proc/mem/performance test single core ---
    Med 229.4 MB/s, Min 214.3  MB/s, Dmin 6.58 Max 239.6  MB/s, Dmax 4.43 Span 11.01
    --- proc/mem/performance test multi-core ---
    Med 1060 MB/s,  Min 1040 MB/s, Dmin 1.89, Max 1100  MB/s, Dmax 3.77, Span 5.66
    
    --- disk test ---
    Seq.Wr. Med 619.4 Min 579.6 Dmin 6.42 Max 628.0 Dmax 1.39 Span 7.81
    Rnd.Wr. Med 214.4 Min 205.5 Dmin 4.14 Max 235.1 Dmax 9.66 Span 13.79
    Seq.Rd. Med 1196  Min 1118  Dmin 6.52 Max 1221  Dmax 2.09 Span 8.61
    Rnd.Rd. Med 933.9 Min 927.4 Dmin 0.69 Max 1000  Dmax 7.08 Span 7.77
    
    --- network test ---
    OK_LON Med **57,4** Min 55,3 Dmin 3,7 Max 58,6 Dmax 2,1 Span 5,7
    AU_MEL Med **22,0** Min 21,6 Dmin 1,6 Max 22,3 Dmax 1,6 Span 3,2
    IN_CHN Med **4,1**  Min 3,3  Dmin 18,5 Max 4,5 Dmax 11,1 Span 29,6
    SG_SGP Med **22,1** Min 21,6 Dmin 2,0 Max 22,6 Dmax 2,5 Span 4,5
    DE_FRA Med **51,8** Min 48,7 Dmin 6,0 Max 53,9 Dmax 4,1 Span 10,0
    IT_MIL Med **46,4** Min 46,0 Dmin 0,8 Max 48,6 Dmax 4,9 Span 5,6
    FR_PAR Med **57,2** Min 52,5 Dmin 8,1 Max 58,1 Dmax 1,7 Span 9,8
    RU_MOS Med **42,2** Min 42,0 Dmin 0,5 Max 43,1 Dmax 2.1 Span 2.6
    BR_SAO Med **1.1**  Min 1.0  Dmin 7.6 Max 1.1  Dmax 4.8 Span 12.4
    US_DAL Med **73.8** Min 67.7 Dmin 8.3 Max 74.5 Dmax 0.9 Span 9.2
    US_SJC Med **46.0** Min 44.9 Dmin 2.4 Max 52.3 Dmax 13.7 Span 16.1
    US_WDC Med **95.6** Min 94.7 Dmin 0.9 Max 96.4 Dmax 0.9 Span 5.0
    JP_TOK Med **22.7** Min 22.7 Dmin 0.0 Max 28.4 Dmax 25.1 Span 25.1
    RO_BUC Med **40.8** Min 39.6 Dmin 2.8 Max 41.0 Dmax 0.6 Span 3.4
    GR_UNK Med **35.1** Min 33.8 Dmin 3.6 Max 35.6 Dmax 1.6 Span 5.1
    NO_OSL Med **48.2** Min 47.9 Dmin 0.5 Max 50.7 Dmax 5.3 Span 5.8
    (Overall Span: 5.8)
    

    First let me note that I would normally not publish results and a review so quickly after just half a day of testing. And I will continue my testing until tomorrow evening (european time) if the provider allows it.
    However, this VPS is so stable that I'm not expecting any slips worth mentioning. Kudos, well done @Hani !

    Let me note something else: I made those tests under FreeBSD. Why? Because when someone noted that they achieved much better results under Linux I of course investigated. After all, I'm looking for solid results. In fact, I even extended my benchmark software to allow for more tunable parameters to allow for digging deeper into a system. What I found was that the reason for Linux' better disk results is simply aggressive caching, or in other words: Linux shows you impressive results but those results aren't stable; they break down under heavy load or when memory gets tight. As my priority is to find the real performance of a system, I'm not interested in nice numbers but in reliable and true real world numbers. Worded differently it seems that Linux answers the question "What's the best you can expect from your disks?" which I'm not interested in, so I stuck with FreeBSD because what some consider a disadvantage actually is an advantage for my testing/benchmarking (but likely not for a newbie with a LAMP stack and light load).

    CPU and memory

    A Xeon 26xx is a decent processor for a server, even older versions (in this case v.2), 2.6 GHz (3.4 GHz turbo) isn't screaming fast but really decent and damn good enough for most users and anyway cache is more important than clock rate for real world performance and the 26xx are quite well equipped. 20 MB L3 cache are good enough to keep the L1 and L2 cache lines well filled.
    As for the memory it's what they say it is (6 GB) and it's DDR3 which is slower than current DDR4 but one must see the whole line from memory to the L1 caches and the 26xx performs nicely. I've seen quite some more current processors who don't do much better overall. Also very positive: Stability. Look at the "span" value! Even in single core mode it's around just 10. Very nice.

    Disks

    I'll keep that one short and dry. The performance isn't great (for an NVMe). One would expect considerably better results from NVMe disks. In the tested system the performance is more like what one would expect from a nice SSD. So, if you want a VPS with real NVMe performance this one is not for you.

    Network

    What I like most is the stability. I've rarely seen such a stable network in the low end segment. Yes, there are some outliers, e.g. Brazil, but a network that offers about the same connectivity and performance to San Jose at the west coast as to Milano, Italy is remarkable - and most of that with sturdy consistence. Very nice!
    I don't know whether ServaRica do have more serious offers for businesses but if I happened to need my (presumably business) web site very well reachable and quite fast at that for customers in both the USA and Europe then ServaRica would be one place to look for that. I'm impressed.
    The downside is that locations outside of the "big business" grid like Sao Paolo, BR or Chennai, IN show really poor results.

    Other/General

    The IP of the VPS I tested is actually in the same /24 as the test IP provided above by ServaRica. I noted that positively; no tricking, just an IP from the really used IP range. Nice.
    As for Xen, I don't care. Xen is OK and the performance is about the same as KVM. Each one has its (mostly minor) advantages and disadvantages but I never had problems with Xen.

    Finally, the price. 7$ for 4 (decent!) vCores, 6 GB memory, and an 80 GB "fast SSD like" disk, along with very nice connectivity is a very attractive deal. One caveat that I want to mention is that the VPS seems to be limited at 100 Mb/s. For most users that shouldn't be a problem at all but if you really need massive bandwidth, then this VPS might not be for you. I personally like it very much though as I know that about 98% of VPSs hardly ever come even close to 100 Mb/s; in fact, I had dedis that were limited at 100 Mb/s and I had no problems with that.

    TL;DR A really sweet deal, especially when you need to be very well - and fast - reachable from both the whole USA and Europe (even Moscow and Bucarest show nice results) and with, let's say, a very nice SSD. Another big plus: That VPS is stable and not the oversold crap one often encounters. Recommended.

Sign In or Register to comment.