Page 7 of 8 FirstFirst ... 5678 LastLast
Results 61 to 70 of 78
 2 Attachment(s)    

Thread: Swvps Very Bad

  1. #61
    Member
    Join Date
    Jan 2009
    Location
    Austria
    Posts
    325

    Default

    Quote Originally Posted by BarrySDCA View Post
    Hi Bill,

    The latency you are concerned with is between the VPS and the broker. The latency between you and the vps is of little concern. The robots that are doing the trading are running in the vps, and you're only sending keyboard/mouse clicks, etc.. At worse, this will cause a little lag in painting your remote desktop screen, but even that can be optimized pretty good. I see this all the time with subscribers from the southern part of Africa. they usually use VNC to connect to get the best remote desktop performance, all while enjoying a trading terminal 16ms away from IBFX. VNC is another remote desktop client software for use in accessing the vps. No real need within the USA to use anthing but MS RDP, which you are probably already using.

    With the least latency between the vps and the broker, the MT4 terminals will enjoy all the benefits.

    In the interest of full disclosure, I am the owner of Commercial Network Services. I was pointed to this thread by a subscriber and could not help but post.

    235ms is quite high. since you are also experiencing high latency to IBFX from your house, I must assume you are also on the east coast. vpsland is based in Atlanta, and so the problem can be on the backbone they and your ISP are using to reach IBFX in Salt Lake City (just an educated guess). With your vps closer to ibfx, you would be immune from this.

    our san diego data center reaches IBFX and back in 16ms (14X faster than what you are doing now). Our New York City data center (due to open March 27) gets to FXDD in 1ms (49 times faster than one you are doing now). this is why vpsland can not possibly have good latency to ibfx...they are in Atlanta, so you immediately have the overhead (and potential points of trouble as you are seeing now) of the x-country IP transit.

    Latency is the physical delay caused by the electrical signals passing through the various copper and fiber infrastructure that make up the core of communications networks. We measure it as the time it takes for a packet of information to reach your broker and come back. Ultra-low latency trading offers tremendous advantages for the trader because price quotes are received and orders entered before most other traders receive the same quote (usually several times over). This is most valuable during periods of volatile market activity when prices change quickly and latency may affect pricing. Low latency also offers the trader more reliable connectivity to the broker because there are fewer hops on the Internet for potential problems to impact the data transmission, requiring the data packet to be sent again at the expense of time. According to Wikipedia, A 1-millisecond advantage in trading applications can be worth $100 million a year.

    Here is our current latency chart. We will be adding another data center in the UK late april/early may.

    Latency to popular brokers from our Trader's Desktop - a VPS for traders - Technical Support
    Barry, Bill

    thank you for informations, but please post your statement in the commercial section.

    Well, Barry's Commercial Network Service is not an ISP, he is located in San Diego, CA and uses the ip range (Class C Net/ 207.158.39.0 - 207.158.39.255) from:

    OrgName: American Internet Services, LLC.
    OrgID: AMERI-504
    Address: 9305 Lightwave Ave.
    City: San Diego
    StateProv: CA
    PostalCode: 92123
    Country: US

    NetRange: 207.158.0.0 - 207.158.63.255
    CIDR: 207.158.0.0/18
    OriginAS: AS6130
    NetName: AIS-WEST2
    NetHandle: NET-207-158-0-0-1
    Parent: NET-207-0-0-0-0
    NetType: Direct Allocation
    NameServer: NS1.AMERICANIS.NET
    NameServer: NS2.AMERICANIS.NET
    Comment: ADDRESSES WITHIN THIS BLOCK ARE NON-PORTABLE.
    RegDate: 1996-06-22
    Updated: 2008-12-10

    Barry's Service is registrated and uses allocated ip range since:

    NetRange: 207.158.39.0 - 207.158.39.255
    CIDR: 207.158.39.0/24
    OriginAS: AS6130
    NetName: CNS-NETBLK-1
    NetHandle: NET-207-158-39-0-1
    Parent: NET-207-158-0-0-1
    NetType: Reassigned
    NameServer: NS1.COMMERCIALNETWORKSERVICES.NET
    NameServer: NS2.COMMERCIALNETWORKSERVICES.NET
    Comment:
    RegDate: 2008-04-24
    Updated: 2008-04-24

    so-10-0.hsa2.sandiego1.level3.net [4.68.113.42] (Backbone ISP)
    american-in.hsa2.sandiego1.level3.net [64.158.218.30] (Level3-American Inteface, American is a customer from Level3)
    sdtc.ar02.g8-10.americanis.net [206.251.233.50] (Exit Router American)
    cns-adnc.commercialnetworkservices.net [209.216.251.26] (American-CNS, CNS is a customer from American)
    vz5.commercialnetworkservices.net [207.158.39.231] (CNS)

    this is just for info, everyone should think about it, whatever he/she want.
    Therefore, statements about other networks, is really dangerous...

    Ixbone (a real ISP :-))
    CCIE#8901 (Ripe AS# 16195)

  2. #62

    Default

    Barry,
    What are your ping stats for FXDD from your place now?? And....IBFX has started working again at VPSLAND...Why was I able to ping IBFX from VPSLAND but have NO CONNECTION to them from MT4?? What are the possible reasons....IBFX MT4 was not able to connect....although FXDD was working fine....?? (It's NOT a MT4 issue as it automatically logs on....) I will consider trying CNS..(the 24/7 3 min live trader assist is worth A LOT)
    Thanks BillR

  3. #63
    Junior Member
    Join Date
    Mar 2009
    Posts
    7

    Default

    Hi Bill,

    From our San Diego data center, FXDD is average 85ms.

    from NYC it is currently 1.71ms

    here are the relevant pings, just run 10:21PM Monday March 16 PDT:

    from san diego data center to fxdd:

    Pinging 129.250.192.70 with 32 bytes of data:

    Reply from 129.250.192.70: bytes=32 time=86ms TTL=54
    Reply from 129.250.192.70: bytes=32 time=87ms TTL=54
    Reply from 129.250.192.70: bytes=32 time=87ms TTL=54
    Reply from 129.250.192.70: bytes=32 time=83ms TTL=54

    Ping statistics for 129.250.192.70:
    Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
    Approximate round trip times in milli-seconds:
    Minimum = 83ms, Maximum = 87ms, Average = 85ms

    From our NYC data center to FXDD:

    $ ping 129.250.192.70
    PING 129.250.192.70 (129.250.192.70): 56 data bytes
    64 bytes from 129.250.192.70: icmp_seq=0 ttl=59 time=1.218 ms
    64 bytes from 129.250.192.70: icmp_seq=1 ttl=59 time=1.143 ms
    64 bytes from 129.250.192.70: icmp_seq=2 ttl=59 time=1.205 ms
    64 bytes from 129.250.192.70: icmp_seq=3 ttl=59 time=1.144 ms
    ^C
    --- 129.250.192.70 ping statistics ---
    4 packets transmitted, 4 packets received, 0.0% packet loss
    round-trip min/avg/max/stddev = 1.143/1.177/1.218/0.034 ms
    $

    We have not had any reported trouble to any broker. We monitor all major brokers and all have been green since the markets opened. anything I say would purely be speculation.

    ixbone: our san diego facilities are located in the American Internet data center, as I have already described above. This is the premiere facility in San Diego. Our cage is a stone throws away from several major companies and even other ISP's. This is very helpful because we can cross-connect with other ISP's very easily and not even have to send our IP traffic over big backbones - that performance routing and latency thing again. We actually ran our own data center back in 1995 thru ~2001. Eventually as the Internet grew, we needed to be in a carrier hotel to get the best connectivity. As an ISP, I'm sure you are doing the same thing. I believe you are misinterpreting the ARIN records. Our ARIN handle is CNS-120. You can read more about both data centers at Commercial Network Services - Data Center

    I certainly did not mean to hijack the thread. I was just trying to give someone an idea of the source of a problem he was experiencing. Since we are the only provider that publically posts our latency to any broker, it was the only chart I had to offer him some comparison to help identify the problem. And since you had already mentioned that you owned an ISP, I thought we were among professionals already and that it was ok to say who I was and post relevant links to the information I was trying to convey. I hope I did not offend anyone. I think I will go back to my day job now.

    Regards,

  4. #64
    Member
    Join Date
    Jan 2009
    Location
    Austria
    Posts
    325

    Default

    Quote Originally Posted by BarrySDCA View Post
    Hi Bill,

    From our San Diego data center, FXDD is average 85ms.

    from NYC it is currently 1.71ms

    here are the relevant pings, just run 10:21PM Monday March 16 PDT:

    from san diego data center to fxdd:

    Pinging 129.250.192.70 with 32 bytes of data:

    Reply from 129.250.192.70: bytes=32 time=86ms TTL=54
    Reply from 129.250.192.70: bytes=32 time=87ms TTL=54
    Reply from 129.250.192.70: bytes=32 time=87ms TTL=54
    Reply from 129.250.192.70: bytes=32 time=83ms TTL=54

    Ping statistics for 129.250.192.70:
    Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
    Approximate round trip times in milli-seconds:
    Minimum = 83ms, Maximum = 87ms, Average = 85ms

    From our NYC data center to FXDD:

    $ ping 129.250.192.70
    PING 129.250.192.70 (129.250.192.70): 56 data bytes
    64 bytes from 129.250.192.70: icmp_seq=0 ttl=59 time=1.218 ms
    64 bytes from 129.250.192.70: icmp_seq=1 ttl=59 time=1.143 ms
    64 bytes from 129.250.192.70: icmp_seq=2 ttl=59 time=1.205 ms
    64 bytes from 129.250.192.70: icmp_seq=3 ttl=59 time=1.144 ms
    ^C
    --- 129.250.192.70 ping statistics ---
    4 packets transmitted, 4 packets received, 0.0% packet loss
    round-trip min/avg/max/stddev = 1.143/1.177/1.218/0.034 ms
    $

    We have not had any reported trouble to any broker. We monitor all major brokers and all have been green since the markets opened. anything I say would purely be speculation.

    ixbone: our san diego facilities are located in the American Internet data center, as I have already described above. This is the premiere facility in San Diego. Our cage is a stone throws away from several major companies and even other ISP's. This is very helpful because we can cross-connect with other ISP's very easily and not even have to send our IP traffic over big backbones - that performance routing and latency thing again. We actually ran our own data center back in 1995 thru ~2001. Eventually as the Internet grew, we needed to be in a carrier hotel to get the best connectivity. As an ISP, I'm sure you are doing the same thing. I believe you are misinterpreting the ARIN records. Our ARIN handle is CNS-120. You can read more about both data centers at Commercial Network Services - Data Center

    I certainly did not mean to hijack the thread. I was just trying to give someone an idea of the source of a problem he was experiencing. Since we are the only provider that publically posts our latency to any broker, it was the only chart I had to offer him some comparison to help identify the problem. And since you had already mentioned that you owned an ISP, I thought we were among professionals already and that it was ok to say who I was and post relevant links to the information I was trying to convey. I hope I did not offend anyone. I think I will go back to my day job now.

    Regards,
    Hi Barry,
    everthing is right, i noticed only you are not an real ISP. Surely, you own some servers and offer hosting and colocation services, but not the networks, thats the difference.
    And yes, you can have a latency of 2msec, if your underlying real isp is next door and a round trip time of 85msec across us is not the best. Interconnected real isp's have a average round trip time of 55msec across us, you must measure from one exit point to the other exit point.
    And a arin customer registration name, is always given from the the real isp to his customers, i make the same with my customers too.
    Thats why i said, you should make your hidden ads, public in the commercial section. Not more, not less.

    IX

  5. #65
    Junior Member
    Join Date
    Mar 2009
    Posts
    7

    Default

    IX: You truly do not know anything about how IP space works. Did you know that much of Verizonís IP space is listed under Microsoft? Verizon is one of the largest ISPís in the country. Does this mean they are not an ISP? Certainly not. When it comes to getting more IP's, I can either get them in a day or two from the datacenter Ė or buy them from any other 3rd party - who will delegate them to us as you have seen, or I can wait weeks if I want my name on top of the pretty ARIN form. I realllly don't care where our name is or even claiming we are an ISP, I do care about delivering IP's and good service quickly to our subscribers. As you know, we are in a big IPv4 shortage and I will take them where I can get them!

    #2: I never said we are an ISP, we are an ASP. I trust you know what that is.

    Additionally, you are so happy to name yourself an ISP. I believe you are doing yourself a disservice. I would never want to host a vps with a single "ISP". It's best to be centrally located in a major datacenter so you have MANY IP transit carriers ISP's to route your traffic. A single ISP will not have any alternate routes in the event of a problem like Bill had earlier. You like to sing "I am an ISP, I am an ISP", but what I hear is "my customers only have one choice of IP transit providers".

    Do you truly expect people to compare you to the REAL ISP's that we utilize? They are: Level3, AT&T, Sprint, UUNet, Qwest, XO, Yipes/Wiltel, Verio, mzima.

    Sir, with all due respect, putting your name on an ARIN form and buying bandwidth from an ISP does not make you an ISP. Nobody wants to purchase IP transit bandwidth across your bedroom. The companies above are ISP's. You purchase bandwidth from onstage.net and have registered your name with the ARIN record. This does not make you an ISP, and if you still insist it does then it does not make you a very diverse or well connected one that I would want to advertise, let alone host my platforms in. What would customers do if onstage.net experienced problems? As an ISP, Iím sure you have peering agreements with other providers? I ran all the looking glass traceroutes that I knew of and I could not find any alternate routes.

    The difference between the two datacenters ping times is because the San Diego datacenter must travel x-country to reach FXDD, who is located in New York. Our datacenters are placed on both coasts for very specific reasons, that Iím sure as an ďISPĒ you will fully understand. How many datacenters are you in? I'm sure you are in a few in order to really be an ISP....got to peer somewhere, right?

    You know just enough to be dangerous.

    Respectfully,

    Barry Bahrami
    Commercial Network Services

  6. #66
    Member
    Join Date
    Jan 2009
    Location
    Austria
    Posts
    325

    Default

    Quote Originally Posted by BarrySDCA View Post
    IX: You truly do not know anything about how IP space works. Did you know that much of Verizonís IP space is listed under Microsoft? Verizon is one of the largest ISPís in the country. Does this mean they are not an ISP? Certainly not. When it comes to getting more IP's, I can either get them in a day or two from the datacenter Ė or buy them from any other 3rd party - who will delegate them to us as you have seen, or I can wait weeks if I want my name on top of the pretty ARIN form. I realllly don't care where our name is or even claiming we are an ISP, I do care about delivering IP's and good service quickly to our subscribers. As you know, we are in a big IPv4 shortage and I will take them where I can get them!

    #2: I never said we are an ISP, we are an ASP. I trust you know what that is.

    Additionally, you are so happy to name yourself an ISP. I believe you are doing yourself a disservice. I would never want to host a vps with a single "ISP". It's best to be centrally located in a major datacenter so you have MANY IP transit carriers ISP's to route your traffic. A single ISP will not have any alternate routes in the event of a problem like Bill had earlier. You like to sing "I am an ISP, I am an ISP", but what I hear is "my customers only have one choice of IP transit providers".

    Do you truly expect people to compare you to the REAL ISP's that we utilize? They are: Level3, AT&T, Sprint, UUNet, Qwest, XO, Yipes/Wiltel, Verio, mzima.

    Sir, with all due respect, putting your name on an ARIN form and buying bandwidth from an ISP does not make you an ISP. Nobody wants to purchase IP transit bandwidth across your bedroom. The companies above are ISP's. You purchase bandwidth from onstage.net and have registered your name with the ARIN record. This does not make you an ISP, and if you still insist it does then it does not make you a very diverse or well connected one that I would want to advertise, let alone host my platforms in. What would customers do if onstage.net experienced problems? As an ISP, Iím sure you have peering agreements with other providers? I ran all the looking glass traceroutes that I knew of and I could not find any alternate routes.

    The difference between the two datacenters ping times is because the San Diego datacenter must travel x-country to reach FXDD, who is located in New York. Our datacenters are placed on both coasts for very specific reasons, that Iím sure as an ďISPĒ you will fully understand. How many datacenters are you in? I'm sure you are in a few in order to really be an ISP....got to peer somewhere, right?

    You know just enough to be dangerous.

    Respectfully,

    Barry Bahrami
    Commercial Network Services
    Well thats the difference...ASP and ISP...i'm connected to Level3, Cogent and Telia, each multi gig lines and additional i'm presented at an national peering point (VIX), where we peer with over 100 real isp's. And no i'm not an isp like Level3, but i'm an full service business provider, which Level3 is not! Don't compare apples with oranges

    To give you the last info to check out, how is an ISP and an ASP like you: My Autonomous System Number is 16195, you can check it out at Ripe (ripe.net) the european registry, similar to Arin.

    peace and regards

    IX

  7. #67
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    After few weeks with EuroVPS and VPSland, that's what I thought : other service, other problems (server down without any reason, hacker intrusion, etc... ).

    Note with SWVPS, I have got two very bad weeks after 6-7 excellent months.

  8. #68
    Junior Member
    Join Date
    Jun 2009
    Posts
    3

    Default

    Quote Originally Posted by funyoo View Post
    After few weeks with EuroVPS and VPSland, that's what I thought : other service, other problems (server down without any reason, hacker intrusion, etc... ).

    Note with SWVPS, I have got two very bad weeks after 6-7 excellent months.


    HI FUNYOO,
    WHICH SWVPS PACKAGE DO YOU USE? THANKS

  9. #69
    Administrator funyoo's Avatar
    Join Date
    Sep 2008
    Posts
    7,003

    Default

    Quote Originally Posted by crastamaya View Post
    HI FUNYOO,
    WHICH SWVPS PACKAGE DO YOU USE? THANKS
    Hi crastamaya,

    WVPS-1

  10. #70

    Default

    Just a heads up..... Commercial Networks Services..CNS has some very serious problems. Barry although a very cordial guy....has a very bad server....During the last months their server has been locked up/ unavailable on numerous occasions...had I not been monitoring my accounts the losses would have been HUGE!!. The customer service has been non existent....Stay AWAY.....It has been NON STOP BS....... Just a very small sample of what to expect below....I could fill pages with their excuses......

    //----------------------------------------------------
    We trapped the error. Appears to be RAM.

    Machine Check Event reported is a fatal memory hierarchy error.
    Trasaction Type: 1
    Memory Hierarchy Level: 2
    Request Type: 0
    Address: 0

    We are taking the node down tomorrow to replace the CPU & RAM (usually done as a pair in production machines).


    Please let me know if you have any questions.

    Thank you,

    Barry Bahrami


    //----------------------------------------------------


    I believe the problem was trapped. the tech is looking at it now.
    Please let me know if you have any questions.

    Thank you,

    Barry Bahrami
    Commercial Network Services


    //---------------------------------------------------------------------------------------------------------------------

    Hello,

    This node has experienced a stop error. We are investigating the trouble now and will keep you updated by service advisory.

    Please let me know if you have any questions.

    Thank you,

    Barry Bahrami
    Commercial Network Services
    //-------------------------------------------..


    We experienced an IP problem in san diego due to abuse from the inside (in a subscribed VE). The offending VE has been identified and shut down. Please let us know if you have any issues.


    Please let me know if you have any questions.

    Thank you,

    Barry Bahrami
    Commercial Network Services
    Last edited by BillR; 06-25-2009 at 04:13.

Tags for this Thread

100, advisor, alpari, alpari uk, automatic, average, best forex expert, broker, brokers, change, check it out, closing, comparison, crossover, demo, eas, expert, expert advisor, expert advisors, forex, forex accounts, forex blog, forex expert advisor, forex vps, free, free ea, guide, high, home, how to, index, japan, live, low, manual, metatrader, mt4, nzd, offer, online, profit, real, robots, sales, sample, server, short term, signals, spread, stop, support, system, systems, test, time, trader, trading, trailing, usd

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •