Howdy, Stranger!

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


Black Friday Exclusive NVMe + Block Storage (BBox) KVM Double RAMs , CPU & Free Credit !!! - Page 6
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.

Black Friday Exclusive NVMe + Block Storage (BBox) KVM Double RAMs , CPU & Free Credit !!!

12346»

Comments

  • letboxletbox Member, Patron Provider
    edited November 2018

    @andiklive said:
    @key900 any 4Gb deals on this BF?

    Wait, i will give you some today or tomorrow.

    Thanked by 1andiklive
  • I reinstalled the VM from scratch and am currently getting the following with virtio/raw block storage:

    # dd if=/dev/zero of=/data/x bs=1G count=1 oflag=dsync
    1+0 records in
    1+0 records out
    1073741824 bytes (1.1 GB) copied, 19.3351 s, 55.5 MB/s
    

    Still seems slow considering 100 MB/s was reported above, but not sure how much better to expect.

  • How much are you paying? What service are you expecting?

  • tetechtetech Member
    edited November 2018

    CyberMonday said: How much are you paying? What service are you expecting?

    Put it like this, BuyVM and LETBox have the same list price ($5/TB) for block storage but with the former I am getting 452 MB/s and the latter it is currently 55 MB/s. I'm expecting something better than an order of magnitude difference.

    Maybe that isn't reasonable, but in that case I'd be grateful if @key900 could tell me what is a reasonable expectation. Especially since I'm only achieving half what others reported in this thread.

  • Just wanted to post a benchmark :smile:

    System Info
    -----------
    Processor   : Common KVM processor
    CPU Cores   : 1
    Frequency   : 2599.998 MHz
    Memory      : 1993 MB
    Swap        :  MB
    Uptime      : 2 days, 11:55,
    
    OS      : Ubuntu 18.04.1 LTS
    Arch        : x86_64 (64 Bit)
    Kernel      : 4.15.0-39-generic
    Hostname    : wtfcook
    
    
    Speedtest (IPv4 only)
    ---------------------
    Your public IPv4 is ****
    
    Location        Provider    Speed
    CDN         Cachefly    57.8MB/s
    
    Atlanta, GA, US     Coloat      19.5MB/s 
    Dallas, TX, US      Softlayer   42.8MB/s 
    Seattle, WA, US     Softlayer   45.4MB/s 
    San Jose, CA, US    Softlayer   40.8MB/s 
    Washington, DC, US  Softlayer   13.0MB/s 
    
    Tokyo, Japan        Linode      5.54MB/s 
    Singapore       Softlayer   8.90MB/s 
    
    Rotterdam, Netherlands  id3.net     5.50MB/s
    Haarlem, Netherlands    Leaseweb    12.9MB/s 
    
    
    Disk Speed
    ----------
    I/O (1st run)   : 136 MB/s
    I/O (2nd run)   : 649 MB/s
    I/O (3rd run)   : 714 MB/s
    Average I/O : 499.667 MB/s
    
  • Damn, hopefully, the io speed is due to people benchmarking simultaneously

  • letboxletbox Member, Patron Provider

    @tetech said:
    I reinstalled the VM from scratch and am currently getting the following with virtio/raw block storage:

    # dd if=/dev/zero of=/data/x bs=1G count=1 oflag=dsync
    1+0 records in
    1+0 records out
    1073741824 bytes (1.1 GB) copied, 19.3351 s, 55.5 MB/s
    

    Still seems slow considering 100 MB/s was reported above, but not sure how much better to expect.

    Why not just using different way to test!

    dd if=/dev/zero of=/dev/xxx bs=1M count=1K maybe?

    Thanked by 1angstrom
  • letboxletbox Member, Patron Provider

    @tetech said:

    CyberMonday said: How much are you paying? What service are you expecting?

    Put it like this, BuyVM and LETBox have the same list price ($5/TB) for block storage but with the former I am getting 452 MB/s and the latter it is currently 55 MB/s. I'm expecting something better than an order of magnitude difference.

    Maybe that isn't reasonable, but in that case I'd be grateful if @key900 could tell me what is a reasonable expectation. Especially since I'm only achieving half what others reported in this thread.

    Do you feel that you paid same as BuyVM charge you? You have 1 GB of RAM free and the Block Storage free as well not sure how do Compare between both offers

    Thanked by 1vimalware
  • key900 said: Why not just using different way to test!

    dd if=/dev/zero of=/dev/xxx bs=1M count=1K maybe?

    Yeah, I did already test it that way too, and it made no difference. The positive news is that both ways are a bit better right now:

    # dd if=/dev/zero of=/data/x bs=1M count=1K
    1024+0 records in
    1024+0 records out
    1073741824 bytes (1.1 GB) copied, 9.38624 s, 114 MB/s
    # umount /data
    # dd if=/dev/zero of=/dev/vdb1 bs=1M count=1K
    1024+0 records in
    1024+0 records out
    1073741824 bytes (1.1 GB) copied, 10.4561 s, 103 MB/s
    
  • key900 said: Do you feel that you paid same as BuyVM charge you? You have 1 GB of RAM free and the Block Storage free as well not sure how do Compare between both offers

    I'm not looking to do a strict comparison between plans, just asking what is a reasonable performance to expect from you. If you tell me "don't expect to get above 100 MB/s on block storage", then OK, that is an answer and I won't spend any more time investigating it.

    If you tell me "you should be getting 200 MB/s", then clearly I'm not, and would want to figure out what is wrong.

  • i still don't get it.... do we get both NVMe and block storage
    i.e. 20 + 256 = 276GB or only 20GB for pack 1?

  • @seenu said:
    i still don't get it.... do we get both NVMe and block storage
    i.e. 20 + 256 = 276GB or only 20GB for pack 1?

    20 + 256

    Thanked by 1seenu
  • @tetech said:

    I'm not looking to do a strict comparison between plans, just asking what is a reasonable performance to expect from you. If you tell me "don't expect to get above 100 MB/s on block storage", then OK, that is an answer and I won't spend any more time investigating it.

    If you tell me "you should be getting 200 MB/s", then clearly I'm not, and would want to figure out what is wrong.

    Block size of writes matters. On a storage array with a lower amount of available iops (like a hdd mirror) using a 1mb or 4mb block size in your 'application' really improves throughput. (your 55MB/s translates to approx 880-900 iops with a 64KB block size)

    So if iops was limited to 200,you'd be seeing single HDD level of write performance. (150-175)

    For Data involving small block/record sizes like databases(4k/8k), storage target should be nvme/ssd arrays.

  • tetechtetech Member
    edited November 2018

    @vimalware said:
    Block size of writes matters. On a storage array with a lower amount of available iops (like a hdd mirror) using a 1mb or 4mb block size in your 'application' really improves throughput. (your 55MB/s translates to approx 880-900 iops with a 64KB block size)

    So if iops was limited to 200,you'd be seeing single HDD level of write performance. (150-175)

    For Data involving small block/record sizes like databases(4k/8k), storage target should be nvme/ssd arrays.

    Thanks for taking the time to write. I guess I shouldn't have been as fixated on a single metric like the dd test.

    What prompted this is I set up the new VM as a DB replication slave and found it to be noticeably slower than any of my other 4 slaves. And they include a VM with spinning disks on an el cheapo $1/month OVZ plan.

    Of course I was a bit bummed to get this performance after signing up for a year with no refund. But oh well, maybe I should have done more research. Live and learn.

    Thanked by 1vimalware
  • letboxletbox Member, Patron Provider
    edited November 2018

    @tetech said:

    key900 said: Do you feel that you paid same as BuyVM charge you? You have 1 GB of RAM free and the Block Storage free as well not sure how do Compare between both offers

    I'm not looking to do a strict comparison between plans, just asking what is a reasonable performance to expect from you. If you tell me "don't expect to get above 100 MB/s on block storage", then OK, that is an answer and I won't spend any more time investigating it.

    If you tell me "you should be getting 200 MB/s", then clearly I'm not, and would want to figure out what is wrong.

    As we mentioned BBox are limited to 100MB/s per VM. NVMe can hit up to 800MB/s and it great for 4k writing if what you asking for!

    Thanked by 1vimalware
  • key900 said: As we mentioned BBox are limited to 100MB/s per VM.

    OK, I didn't see that stated anywhere before I ordered, or else I wouldn't have proceeded.

  • I suspect their router is misconfigured seriously.
    For example, their router advertises 5 IPv6 prefixes including quite large one and let users to use any of their IP range.

    @key900
    Please read howto configure your router and implement anti-spoofing countermeasures.

  • letboxletbox Member, Patron Provider

    @cause said:
    I suspect their router is misconfigured seriously.
    For example, their router advertises 5 IPv6 prefixes including quite large one and let users to use any of their IP range.

    @key900
    Please read howto configure your router and implement anti-spoofing countermeasures.

    Are you referring to IPv6 or both?

  • letboxletbox Member, Patron Provider

    @cause said:
    I suspect their router is misconfigured seriously.
    For example, their router advertises 5 IPv6 prefixes including quite large one and let users to use any of their IP range.

    @key900
    Please read howto configure your router and implement anti-spoofing countermeasures.

    Some people were effected like 20 people that are different pool, we have fix some of them and we still working for the rest.

  • @tetech said:

    key900 said: As we mentioned BBox are limited to 100MB/s per VM.

    OK, I didn't see that stated anywhere before I ordered, or else I wouldn't have proceeded.

    @key900 I think this is worth mentioning in your offer copy text in future

    Else you'll be dealing with lots of disappointed signups who saw the 1.1GB/s dd write test results from your previous offer threads.

    They don't always understand that, that is because all benchmark scripts write to user's home or /tmp/ for their measurements. (NVMe block device)

    Thanked by 1tetech
  • letboxletbox Member, Patron Provider

    @vimalware said:

    @tetech said:

    key900 said: As we mentioned BBox are limited to 100MB/s per VM.

    OK, I didn't see that stated anywhere before I ordered, or else I wouldn't have proceeded.

    @key900 I think this is worth mentioning in your offer copy text in future

    Else you'll be dealing with lots of disappointed signups who saw the 1.1GB/s dd write test results from your previous offer threads.

    They don't always understand that, that is because all benchmark scripts write to user's home or /tmp/ for their measurements. (NVMe block device)

    Great, your feedback has been considered, if there is any bug or any feedback i’m happ to see them all and learning of it!

    Thanks you.

  • @key900 said:

    @cause said:
    I suspect their router is misconfigured seriously.
    For example, their router advertises 5 IPv6 prefixes including quite large one and let users to use any of their IP range.

    @key900
    Please read howto configure your router and implement anti-spoofing countermeasures.

    Some people were effected like 20 people that are different pool, we have fix some of them and we still working for the rest.

    Both. IPv6 problem can be verified with rdisc6 eth0 and assign available address space. I cannot verify IPv4 side since it may interfere other users.

  • letboxletbox Member, Patron Provider

    @cause said:

    @key900 said:

    @cause said:
    I suspect their router is misconfigured seriously.
    For example, their router advertises 5 IPv6 prefixes including quite large one and let users to use any of their IP range.

    @key900
    Please read howto configure your router and implement anti-spoofing countermeasures.

    Some people were effected like 20 people that are different pool, we have fix some of them and we still working for the rest.

    Both. IPv6 problem can be verified with rdisc6 eth0 and assign available address space. I cannot verify IPv4 side since it may interfere other users.

    Nothing wrong with IPv4 i can confirm, for the IPv6 we will do check.

  • @key900
    Your cisco router is requesting arp on wrong interface and one can reply to any of them. Is not it wrong?

  • nhutphannhutphan Member
    edited November 2018

    Is double credit still valid? :)

  • @nhutphan said:
    Is double credit still valid? :)

    Still

    Thanked by 1nhutphan
  • When does the free credit drop bruh?

  • @corbpie said:
    When does the free credit drop bruh?

    never

  • letboxletbox Member, Patron Provider

    @sonic said:

    @corbpie said:
    When does the free credit drop bruh?

    never

    All free credit are set from day 1 10x $10 and 10 x $5

    To see the credit Balance

    Client Area >> on the right above Notifications button

Sign In or Register to comment.