Howdy, Stranger!

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


Dacentec network speeds
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.

Dacentec network speeds

Hi guys,

I just bought a bare metal server(Supermicro X3440 8GB 2x2TB SATA Rent Only) from Dacentec a month ago and have been having unstable network speeds. For those who have had or currently own a server with Dacentec, what are your experiences with their network reliability and speeds.

Link to speedtest.net results

freevps bench script

Administrator@otaku ~ $ ./bench.sh CPU model : Intel(R) Xeon(R) CPU X3440 @ 2.53GHz Number of cores : 8 CPU frequency : 2533.000 MHz Total amount of ram : MB Total amount of swap : MB System uptime : Download speed from CacheFly: 7.62MB/s Download speed from Coloat, Atlanta GA: 10.9MB/s Download speed from Softlayer, Dallas, TX: 3.70MB/s Download speed from Linode, Tokyo, JP: 1.22MB/s Download speed from i3d.net, Rotterdam, NL: 336KB/s Download speed from Leaseweb, Haarlem, NL: 3.51MB/s Download speed from Softlayer, Singapore: 745KB/s Download speed from Softlayer, Seattle, WA: 1.98MB/s Download speed from Softlayer, San Jose, CA: 1.57MB/s Download speed from Softlayer, Washington, DC: 2.96MB/s I/O speed : 12.8 MB/s

«1345

Comments

  • @dacentec, Do you know what speeds I should be expecting? When I purchased the server, I was aware it was on a 1Gbps port, but nothing was specified about the network speed.

  • Open a ticket up with them and give them some info, they'll resolve it.

    The network speeds are usually pretty good.

  • tomletomle Member, LIR
    edited July 2015

    Looks really bad, and so does your I/O. I've got two boxes with them, the following is the one that is worse:

    CPU model : Intel(R) Xeon(R) CPU L5420 @ 2.50GHz Number of cores : 8 CPU frequency : 2499.000 MHz Total amount of ram : 15886 MB Total amount of swap : 8063 MB System uptime : 56 days, 7:12, Download speed from CacheFly: 97.0MB/s Download speed from Coloat, Atlanta GA: 21.2MB/s Download speed from Softlayer, Dallas, TX: 39.0MB/s Download speed from Linode, Tokyo, JP: 7.59MB/s Download speed from i3d.net, Rotterdam, NL: 4.60MB/s Download speed from Leaseweb, Haarlem, NL: 20.6MB/s Download speed from Softlayer, Singapore: 8.60MB/s Download speed from Softlayer, Seattle, WA: 19.9MB/s Download speed from Softlayer, San Jose, CA: 27.6MB/s Download speed from Softlayer, Washington, DC: 42.7MB/s I/O speed : 160 MB/s (RAID0)

    This is the better one:

    CPU model : Quad-Core AMD Opteron(tm) Processor 1385 Number of cores : 4 CPU frequency : 2700.000 MHz Total amount of ram : 15949 MB Total amount of swap : 8031 MB System uptime : 56 days, 21:21, Download speed from CacheFly: 93.8MB/s Download speed from Coloat, Atlanta GA: 22.6MB/s Download speed from Softlayer, Dallas, TX: 66.4MB/s Download speed from Linode, Tokyo, JP: 12.6MB/s Download speed from i3d.net, Rotterdam, NL: 19.3MB/s Download speed from Leaseweb, Haarlem, NL: 19.0MB/s Download speed from Softlayer, Singapore: 7.97MB/s Download speed from Softlayer, Seattle, WA: 30.7MB/s Download speed from Softlayer, San Jose, CA: 31.6MB/s Download speed from Softlayer, Washington, DC: 59.8MB/s I/O speed : 84.3 MB/s (RAID1)

  • Thanks @tomle. I opened a ticket and was transferred from the support department to the network department, but ticket responses have been slower than usual. I was just told to run their version of the bench script (http://mirror.dacentec.com/bench.sh). Same results, as expected. Still no response.

  • Also, yeah, my I/O is atrocious. My USB2.0 flash drive is faster than that...

  • FredQcFredQc Member
    Dacentec Server Test
    based on freevps bench.sh
    
    CPU model :  Intel(R) Xeon(R) CPU           L5520  @ 2.27GHz
    Number of cores : 16
    CPU frequency :  2268.000 MHz
    Total amount of ram : 72347 MB
    Total amount of swap : 16383 MB
    System uptime :   21 days, 3:05,        
    
    Listing Memory Modules 
    ...
    
    Listing HAL Info 
    Board Manufacturer: 
    Model: 
    Version: 
    Serial No.: 
    
    
    Disk Listing
    VENDOR  -   MODEL   -    DEVICE  -   MOUNT  -   LABEL  -    SIZE 
    
     ATA - Crucial_CT240M50 - /dev/sdb - Mount-N/A -  l5520:1 -  240GB 
    
     ATA - Crucial_CT240M50 - /dev/sda - Mount-N/A -  l5520:1 -  240GB 
    
    
    USB Disks
    
    
    CDROM Drives
    
    
    Network Tests
    Download speed from CacheFly: 91.8MB/s 
    Download speed from Coloat, Atlanta GA: 20.2MB/s 
    Download speed from Softlayer, Dallas, TX: 49.5MB/s 
    Download speed from Linode, Tokyo, JP: 11.8MB/s 
    Download speed from Leaseweb, Haarlem, NL: 67.7MB/s 
    Download speed from Softlayer, Singapore: 6.67MB/s 
    Download speed from Softlayer, Seattle, WA: 16.0MB/s 
    Download speed from Softlayer, San Jose, CA: 23.7MB/s 
    Download speed from Softlayer, Washington, DC: 51.8MB/s 
    Download speed from Dacentec, Lenoir, NC USA: 100MB/s 
    
    
    Disk IO Test
    I/O speed :  374 MB/s
    
  • Make sure you enable writeback cache on the drives in your server. I remember seeing another thread where @MarkTurner was having issues with one of his customers and the issue with slow IO was that writeback cache wasn't enabled.

    hdparm -W1 /dev/sdX

    Try that and see if your IO improves at least.

    Cheers!

  • @TheLinuxBug. Thanks for the tip! I will make sure I do that if I switch to linux. I'm currently running Windows Server 2012 R2. I ran those benchmark scripts in cygwin which shouldn't make a difference in terms of network speeds. Do you know if writeback cache needs to be enabled for Windows Server?

  • Chris just responded to the ticket.

    Hello

    The unit is rebooting to your OS now. The results of the baremetal test are as follows:

    Network Tests
    Download speed from CacheFly: 99.6MB/s
    Download speed from Coloat, Atlanta GA: 20.4MB/s
    Download speed from Softlayer, Dallas, TX: 37.6MB/s
    Download speed from Linode, Tokyo, JP: 11.5MB/s
    Download speed from Leaseweb, Haarlem, NL: 44.5MB/s
    Download speed from Softlayer, Singapore: 5.05MB/s
    Download speed from Softlayer, Seattle, WA: 16.6MB/s
    Download speed from Softlayer, San Jose, CA: 20.7MB/s
    Download speed from Softlayer, Washington, DC: 14.3MB/s
    Download speed from Dacentec, Lenoir, NC USA: 109MB/s

    Again, these values are within the normal range we would expect to see.

    Thanks
    Chris H.

    I have no idea what could be limiting my downstream within Windows Server. Any theories?

  • VPSSoldiersVPSSoldiers Member
    edited July 2015

    This is one of my servers there.

    CPU model :  Intel(R) Xeon(R) CPU           L5520  @ 2.27GHz
    Number of cores : 16
    CPU frequency :  2266.663 MHz
    Total amount of ram : 72496 MB
    Total amount of swap : 110595 MB  
    Download speed from CacheFly: 91.6MB/s 
    Download speed from Coloat, Atlanta GA: 24.2MB/s 
    Download speed from Softlayer, Dallas, TX: 48.6MB/s 
    Download speed from Linode, Tokyo, JP: 6.60MB/s 
    Download speed from i3d.net, Rotterdam, NL: 2.94MB/s
    Download speed from Leaseweb, Haarlem, NL: 46.6MB/s 
    Download speed from Softlayer, Singapore: 6.87MB/s 
    Download speed from Softlayer, Seattle, WA: 22.1MB/s 
    Download speed from Softlayer, San Jose, CA: 21.5MB/s 
    Download speed from Softlayer, Washington, DC: 3.56MB/s 
    I/O speed :  178 MB/s (RAID10)
    
  • @VPSSoldiers - thats a lot of swap:

    Total amount of swap : 110595 MB

  • Yes.

  • Which is best, dacentec $25 monthly or delimiter $20 monthly dedicated server? Both with 8gb ram.

  • MarkTurnerMarkTurner Member
    edited July 2015

    @XxNisseGamerxX - thats depends on use. Dacentec give more disk (2x2TB vs 500GB), Delimiter has better connectivity and better location.

    Also Delimiter gives you dedicated IPMI/KVM 24x7 rather than on demand (I know thats ironic considering your Java issues). For customers using management tools thats a bonus because they use various ILO tools to reboot their servers, pull management logs, etc

  • sinsin Member

    root@sin:~# benchmark
    CPU model : Quad-Core AMD Opteron(tm) Processor 1381
    Number of cores : 4
    CPU frequency : 2500.000 MHz
    Total amount of ram : 8003 MB
    Total amount of swap : 8095 MB
    System uptime : 3 days, 22:08,
    Download speed from CacheFly: 97.1MB/s
    Download speed from Coloat, Atlanta GA: 24.4MB/s
    Download speed from Softlayer, Dallas, TX: 40.0MB/s
    Download speed from Linode, Tokyo, JP: 9.23MB/s
    Download speed from i3d.net, Rotterdam, NL: 5.93MB/s
    Download speed from Leaseweb, Haarlem, NL: 29.6MB/s
    Download speed from Softlayer, Singapore: 4.88MB/s
    Download speed from Softlayer, Seattle, WA: 21.8MB/s
    Download speed from Softlayer, San Jose, CA: 23.2MB/s
    Download speed from Softlayer, Washington, DC: 56.1MB/s
    I/O speed : 93.7 MB/s

  • dacentecdacentec Member, Host Rep

    MarkTurner said: Delimiter has better connectivity (500GB) and better location.

    What is your AS / Test IP? I don't see it in your ads.

    IPMI keeps coming up as a security issue.
    https://www.us-cert.gov/ncas/alerts/TA13-207A

    Thanked by 2vimalware geekalot
  • And the battle begins.

  • dacentecdacentec Member, Host Rep

    No battle, just curious/interested. Maybe a delimiter customer can check at http://www.whatismyasn.org/

  • dacentec said: No battle, just curious/interested. Maybe a delimiter customer can check at http://www.whatismyasn.org/

    not a customer of them but from previous posts i believe AS7363 is linked to them.

  • qpsqps Member, Host Rep

    MarkTurner said: Delimiter has better connectivity (500GB)

    Are you stating that Delimiter/Yomura/etc has 500 Gbps of IP transit uplink capacity in Atlanta?

  • MarkTurnerMarkTurner Member
    edited July 2015

    qps said: Are you stating that Delimiter/Yomura/etc has 500 Gbps of IP transit uplink capacity in Atlanta?

    LOL that was me writing something too quickly on my phone. I meant to say our server is including 500GB of disk space vs their 2x2TB, the cursor ended up in the wrong place when I wrote it.

    Delimiter has access to 30Gbps of traffic in Yomura's datacentre.

    [Edited: The 500GB reference is now in the right place]

  • dacentec said: What is your AS / Test IP? I don't see it in your ads.

    Delimiter in Atlanta sits under AS7363

  • dacentecdacentec Member, Host Rep
    edited July 2015

    Hard to find delimiter test IP in your ads.
    https://www.peeringdb.com/private/participant_view.php?id=4970
    http://bgp.he.net/AS7363#_asinfo

    Dacentec
    http://bgp.he.net/AS31863
    https://www.peeringdb.com/private/participant_view.php?id=3587

    @markturner your peeringdb says you are up on the Atlanta IX, but I am not seeing you here: https://tie.telx.com/#participants, and I can't trace or ping to you from my 10 gig IX uplink.

    I was able to find qps test ip pretty quick with google. Nice network.

    In case others are wondering where the big ISPs sit:
    http://research.dyn.com/2015/02/bakers-dozen-2014-edition/

  • lmnotranlmnotran Member
    edited July 2015

    Wow, this thread got interesting. Thanks for the help everyone! I decided to take the nuclear route and re-provisioned my server from Windows Server 2012 R2 to Ubuntu Server 15.04. Things seem to be fixed. It's still weird to me how something was limiting my downstream in Windows. There weren't and group policies set up that would affect network.

    mason@otaku:~$ mason@otaku:~$ wget http://mirror.dacentec.com/bench.sh
    mason@otaku:~$ chmod +x bench.sh
    mason@otaku:~$ ./bench.sh
    
    Dacentec Server Test
    based on freevps bench.sh
    
    CPU model :  Intel(R) Xeon(R) CPU           X3440  @ 2.53GHz
    Number of cores : 8
    CPU frequency :  1200.000 MHz
    Total amount of ram : 7975 MB
    Total amount of swap : 91 MB
    System uptime :   8 min,
    
    Listing Memory Modules
    /dev/mem: Permission denied
    
    Listing HAL Info
    
    Linux HAL not found, skipping advanced hardware detection
    
    Network Tests
    Download speed from CacheFly: 105MB/s
    Download speed from Coloat, Atlanta GA: 17.1MB/s
    Download speed from Softlayer, Dallas, TX: 37.7MB/s
    Download speed from Linode, Tokyo, JP: 11.9MB/s
    Download speed from Leaseweb, Haarlem, NL: 60.2MB/s
    Download speed from Softlayer, Singapore: 5.97MB/s
    Download speed from Softlayer, Seattle, WA: 17.1MB/s
    Download speed from Softlayer, San Jose, CA: 18.5MB/s
    Download speed from Softlayer, Washington, DC: 45.5MB/s
    Download speed from Dacentec, Lenoir, NC USA: 109MB/s
    
    Disk IO Test
    I/O speed :  133 MB/s
    

    @dacentec: Just curious, are you one of the Chris's that I've been bugging constantly?

    @MarkTurner: I've been really tempted to grab a server from you guys located in Minneapolis, but the servers that are offered only include a 1TB HDD. Do you guys have any plans to offer any servers more storage oriented, similar to the one I currently have (X3440 8GB 2x2TB SATA)?

  • dacentecdacentec Member, Host Rep

    lmnotran said: @dacentec: Just curious, are you one of the Chris's that I've been bugging constantly?

    No, but those guys are great.

  • lmnotranlmnotran Member
    edited July 2015

    They are! Do they live in the data center and switch off every 12 hours? LOL

  • It's one guy.

    Watch foamy the squirrel tech support 3

    And tech support 2 lmfao

    Thanked by 1doughmanes
  • MarkTurner said: @XxNisseGamerxX - thats depends on use. Dacentec give more disk (2x2TB vs 500GB), Delimiter has better connectivity and better location.

    100Mbps unlimited option yet? :(

  • SadySady Member

    Dacentec's support is fucking awesome. All who are comparing Delimeter/Dacentec, I would suggest you all to give Dacentec a try & you won't regret.

    Thanked by 2gestiondbi RBH
  • zevuszevus Member

    I haven't tested in a month or so, but I had burst speeds up to around 50MB/s (5-10 seconds) w/ something like 15-20MB/s "sustained" (some concurrent FTPs that I cancelled after some 5 minutes or so).

    So considering prices, good speeds. Back in January-Feb, I had higher speeds, but couldn't say if it's just because of new 'neighbors' that are sharing my gigabit port, or if it's slowed down a bit overall.....

    regardless, nothing to complain about w/ current speeds on my $20/mo server

Sign In or Register to comment.