Howdy, Stranger!

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


[Review] Contabo VPS M SSD - Page 2
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.

[Review] Contabo VPS M SSD

2»

Comments

  • I got around 1/3 of the speed of an ovh ssd 1 while using their 2 core offer. My experience has been that their performance is a joke and their Network will randomly go down in small increments throughout the day.

  • deer76deer76 Member

    Posting the full test again, the initial write is always a bit slow. It looks like a spinning disk warm-up but obviously this cannot be the case anymore nowadays :p

    For now I am keeping this, let's see how it goes.

    -------------------------------------------------
     nench.sh v2018.04.14 -- https://git.io/nench.sh
     benchmark timestamp:    2018-05-05 06:26:41 UTC
    -------------------------------------------------
    
    Processor:    Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz
    CPU cores:    4
    Frequency:    2199.998 MHz
    RAM:          11G
    Swap:         13G
    Kernel:       Linux 4.13.0-39-generic x86_64
    
    Disks:
    sda    300G  HDD
    
    CPU: SHA256-hashing 500 MB
        4.735 seconds
    CPU: bzip2-compressing 500 MB
        8.221 seconds
    CPU: AES-encrypting 500 MB
        2.299 seconds
    
    ioping: seek rate
        min/avg/max/mdev = 40 us / 157 us / 12.4 ms / 385 us
    ioping: sequential read speed
        18.9 k requests completed in 5.00 s, 3.83 k iops, 957.9 MiB/s
    
    dd: sequential write speed
        1st run:    169.75 MiB/s
        2nd run:    322.34 MiB/s
        3rd run:    318.53 MiB/s
        average:    270.21 MiB/s
    
    IPv4 speedtests
        your IPv4:    xxxx
    
        Cachefly CDN:         11.49 MiB/s
        Leaseweb (NL):        11.54 MiB/s
        Softlayer DAL (US):   1.87 MiB/s
        Online.net (FR):      11.39 MiB/s
        OVH BHS (CA):         10.04 MiB/s
    
    Thanked by 1Ympker
  • YmpkerYmpker Member
    edited May 2018

    Uptimerobot monitoring a Contabo IP:

    Since 119 hrs, 44 mins (2018-04-30 11:01:59)

    Uptime

    100% (last 24 hours)

    99.93% (last 7 days)

    99.98% (last 30 days)

    I can also share uptimerobot logs worth of some months with you if intetested :)

    Thanked by 1deer76
  • AntKalaAntKala Member

    I just don't understand why you would get a vps over a dedicated server if they are worth the same price. I've rented contabo's largest vps for 27 euro, and any dedi from another company at the same price will outperform it.

  • jackbjackb Member, Host Rep

    @AntKala said:
    I just don't understand why you would get a vps over a dedicated server if they are worth the same price. I've rented contabo's largest vps for 27 euro, and any dedi from another company at the same price will outperform it.

    27EUR dedi might not come with RAID and won't come with dual power supplies. Etc.

    I've no idea if contabo's nodes do - but a large number of VPS providers will have nodes with useful things like that - improving your uptime.

    Thanked by 1pullangcubo
  • AnthonySmithAnthonySmith Member, Patron Provider

    AntKala said: I just don't understand why you would get a vps over a dedicated server if they are worth the same price. I've rented contabo's largest vps for 27 euro, and any dedi from another company at the same price will outperform it.

    I would suggest that unless you got a REALLY good deal, any dedi you get for €27 will not even out perform the VPS in the last bench mark, will certainly have slower disks, much older CPU's be less redundant and probably has no IPMI (VPS has pseudo with ISO options via VNC)

    Maybe I am wrong though?

  • FHRFHR Member, Host Rep

    AnthonySmith said: Maybe I am wrong though?

    You are definitely not wrong. Closest thing for the price is Kimsufi KS-4C – i5-2300, 16GB DDR3, 2TB HDD, 100Mbps port for 23€, which I wouldn't exactly call reliable with AIO motherboards and 1 HDD.

  • FoxelVoxFoxelVox Member
    edited May 2018

    What are you expecting? you pay 9 euro for 12GB ram and 4 pretty good era CPU cores.. OVH gives you WAAAAY less resources to play with. And judging by the benchmarks above, their read/writes and iops are extraordinairy for the price.

  • HarambeHarambe Member, Host Rep

    Ympker said: I can also share uptimerobot logs worth of some months with you if intetested :)

    Would be interested in seeing those if you want to share :)

    Thanked by 1Ympker
  • deer76deer76 Member
    edited May 2018

    @FoxelVox said:
    What are you expecting? you pay 9 euro for 12GB ram and 4 pretty good era CPU cores.. OVH gives you WAAAAY less resources to play with. And judging by the benchmarks above, their read/writes and iops are extraordinairy for the price.

    I confirm that it's a really good deal for the price.

    I am expecting what is advertised, that's the basic of a honest service market. I also have interest for the good providers to grow in a healthy market.

    So far it seems great, I am barely using network so can't comment on that (will add monitoring later) but the vps behaves consistently. I will do another review after a few months, this was the initial impact review.

    The fact you have to ask to remove the IO limitation is kinda weird, I wouldn't have known without reading about it here on LET. Nothing else negative to report so far.

  • edfoxedfox Member

    This is mine, i didn't ask to remove limitations or anything - i just installed stock debian from the sysrescue:


    nench.sh v2018.04.14 -- https://git.io/nench.sh

    benchmark timestamp: 2018-05-06 08:50:04 UTC

    Processor: Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz
    CPU cores: 4
    Frequency: 2199.996 MHz
    RAM: 11Gi
    bash: line 156: swapon: command not found
    Swap: -
    Kernel: Linux 4.15.0-2-amd64 x86_64

    Disks:
    sda 300G HDD

    CPU: SHA256-hashing 500 MB
    6.057 seconds
    CPU: bzip2-compressing 500 MB
    11.107 seconds
    CPU: AES-encrypting 500 MB
    2.315 seconds

    ioping: seek rate
    min/avg/max/mdev = 36.3 us / 85.3 us / 12.2 ms / 218.1 us
    ioping: sequential read speed
    generated 11.7 k requests in 5.00 s, 2.86 GiB, 2.34 k iops, 586.0 MiB/s

    dd: sequential write speed
    1st run: 70.00 MiB/s
    2nd run: 102.04 MiB/s
    3rd run: 102.04 MiB/s
    average: 91.36 MiB/s

    IPv4 speedtests
    your IPv4: 173.249.20.xxxx

    Cachefly CDN:         11.54 MiB/s
    Leaseweb (NL):        11.41 MiB/s
    Softlayer DAL (US):   5.58 MiB/s
    Online.net (FR):      10.99 MiB/s
    OVH BHS (CA):         5.18 MiB/s
    

    No IPv6 connectivity detected


    nench.sh v2018.04.14 -- https://git.io/nench.sh

    benchmark timestamp: 2018-05-06 08:51:58 UTC

    Processor: Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz
    CPU cores: 4
    Frequency: 2199.996 MHz
    RAM: 11Gi
    bash: line 156: swapon: command not found
    Swap: -
    Kernel: Linux 4.15.0-2-amd64 x86_64

    Disks:
    sda 300G HDD

    CPU: SHA256-hashing 500 MB
    4.893 seconds
    CPU: bzip2-compressing 500 MB
    8.205 seconds
    CPU: AES-encrypting 500 MB
    1.951 seconds

    ioping: seek rate
    min/avg/max/mdev = 35.3 us / 79.5 us / 13.5 ms / 171.7 us
    ioping: sequential read speed
    generated 11.7 k requests in 5.00 s, 2.86 GiB, 2.34 k iops, 586.0 MiB/s

    dd: sequential write speed
    1st run: 145.91 MiB/s
    2nd run: 103.00 MiB/s
    3rd run: 103.00 MiB/s
    average: 117.30 MiB/s

    IPv4 speedtests
    your IPv4: 173.249.20.xxxx

    Cachefly CDN:         1 MiB/s
    Leaseweb (NL):        2.06 MiB/s
    Softlayer DAL (US):   1.53 MiB/s
    Online.net (FR):      11.09 MiB/s
    OVH BHS (CA):         6.24 MiB/s
    

    No IPv6 connectivity detected

  • YmpkerYmpker Member

    @Harambe said:

    Ympker said: I can also share uptimerobot logs worth of some months with you if intetested :)

    Would be interested in seeing those if you want to share :)

    PM to remind me. I'm not feeling very well today. Probably tomorrow :)

  • edfoxedfox Member

    Thank you very much for your e-mail.

    Our VPS infrastructure is built for optimal overall performance for all customers. Our benchmarks have shown that with our current default settings, latency is low and IOPS are at a maximum. These two factors are crucial for fast reaction times and high processing speed of typical server applications such as databases.

    We have adjusted the parameters of your VPS so that significantly higher sequential write throughput can now be achieved as well. The change is effective immediately.

    If you have any questions or need help, please do not hesitate to contact us.

    Before:

    dd: sequential write speed
    1st run: 145.00 MiB/s
    2nd run: 101.91 MiB/s
    3rd run: 105.00 MiB/s
    average: 117.30 MiB/s

    After:

    dd: sequential write speed
    1st run: 204.09 MiB/s
    2nd run: 358.58 MiB/s
    3rd run: 148.77 MiB/s
    average: 237.15 MiB/s

  • deer76deer76 Member

    I can still see some "warm-up" issues from time to time:

    root@vps:~# ./bench.sh -io
    Disk Speed
    ----------
    I/O (1st run)   : 3.3 MB/s
    I/O (2nd run)   : 24.1 MB/s
    I/O (3rd run)   : 252 MB/s
    Average I/O : 93.1333 MB/s
    
    root@vps:~# ./bench.sh -io
    Disk Speed
    ----------
    I/O (1st run)   : 335 MB/s
    I/O (2nd run)   : 334 MB/s
    I/O (3rd run)   : 320 MB/s
    Average I/O : 329.667 MB/s
    

    No idea why is that..I find it puzzling

  • deer76deer76 Member
    edited July 2018

    Today it's the first time that I notice disk performance completely thrashed. I can safely say that it has been as reported above in all the past months.

    Now it is so slow that basically bench.sh -io never completes..or it's going to take around 10 minutes since about 5 have already elapsed.

    Does any other Contabo user have similar issues? They might have some problem...let's see what is their reply.

    Edit: issue seems to have been temporary, I will observe what happens in future. It lasted about 1h; now everything seems back to the usual speed:

    Disk Speed


    I/O (1st run) : 284 MB/s
    I/O (2nd run) : 362 MB/s
    I/O (3rd run) : 318 MB/s
    Average I/O : 321.333 MB/s

    Disk Speed


    I/O (1st run) : 344 MB/s
    I/O (2nd run) : 376 MB/s
    I/O (3rd run) : 380 MB/s
    Average I/O : 366.667 MB/s

  • ShazanShazan Member, Host Rep

    Perhaps an array rebuild?

  • @deer76 said:
    The fact you have to ask to remove the IO limitation is kinda weird, I wouldn't have known without reading about it here on LET. Nothing else negative to report so far.

    Do you really need to do that still? I did test my box using Serverscope a few months back and the the speeds were still superb:
    https://serverscope.io/trials/n23R

  • Today noticed again very sluggish performance (like, can't edit a file remotely). Checked the I/O:

    Disk Speed
    ----------
    I/O (1st run)   : 16.1 MB/s
    I/O (2nd run)   : 38.2 MB/s
    I/O (3rd run)   : 291 MB/s
    Average I/O : 115.1 MB/s
    

    Following runs:

    Disk Speed
    ----------
    I/O (1st run)   : 102 MB/s
    I/O (2nd run)   : 156 MB/s
    I/O (3rd run)   : 337 MB/s
    Average I/O : 198.333 MB/s
    
    Disk Speed
    ----------
    I/O (1st run)   : 355 MB/s
    I/O (2nd run)   : 379 MB/s
    I/O (3rd run)   : 372 MB/s
    Average I/O : 368.667 MB/s
    
    Disk Speed
    ----------
    I/O (1st run)   : 167 MB/s
    I/O (2nd run)   : 343 MB/s
    I/O (3rd run)   : 353 MB/s
    Average I/O : 287.667 MB/s
    
    Disk Speed
    ----------
    I/O (1st run)   : 323 MB/s
    I/O (2nd run)   : 347 MB/s
    I/O (3rd run)   : 356 MB/s
    Average I/O : 342 MB/s
    

    @Prime404: Yes, I had to do that once at the beginning. And I cannot explain why sometimes performance becomes completely thrashed, probably a noisy neighbour

  • Again today:

    Disk Speed
    ----------
    I/O (1st run)   : 4.9 MB/s
    I/O (2nd run)   : 153 MB/s
    I/O (3rd run)   : 238 MB/s
    Average I/O : 131.967 MB/s
    

    Then after two runs of the IO benchmark it becomes faster again.

    I am considering switching to a provider that has no such "benchmark detection" in place...

Sign In or Register to comment.