Page 2 of 8 FirstFirst 1234 ... LastLast
Results 11 to 20 of 78
  1. #11
    Untangle Ninja
    Join Date
    Jan 2011
    Posts
    1,327

    Default

    Quote Originally Posted by sky-knight View Post
    Right, and can that unit you have at home keep up with a DOCSIS 3.0 cable line?
    I've never heard of anyone around here with such a fast connection. For whatever reasons, genuinely high-speed internet is just very rare around here. I guess if I ran into anyone with a really fast connection at home, I'd steer them to my more expensive business ut with the intel nics.

  2. #12
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,554

    Default

    You apparently don't live in Phoenix. I have multiple customers, myself included that have 10/100 DOCSIS 3.0 connections.

    I haven't seen a realtek clear 20mb. So if you've got the 50mb DOCSIS 2.0, you're STILL not keeping up. And those connections are everywhere.
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  3. #13
    Untanglit ixeous's Avatar
    Join Date
    Sep 2007
    Posts
    24

    Default

    Quote Originally Posted by YeOldeStonecat View Post
    But what happens in you put a real load on it? Just passing a speed test traffic like it's some Stinksys or Nutgear home grade router doesn't do much for testimony of a WinNIC. Mix in lots of different web traffic, e-mail scanning, QoS being applied to various traffic, bidirectional traffic, in other words...."real world loads of a decent sized business network"...done in layer 7.
    This is an urealistic testing standard. It is excessively difficult to truely simulate 100, 200, or 500 users (x number threads from a single host is not the same) with a variety of traffic (DNS, HTTP, HTTPS, FTP, SMTP, SSH, VPN, XMPP, etc) crossing Untangle all of which have varying packet sizes, to multiple server hosts, with varying file types and varying download sizes while ensuring that the bottleneck remains the Untangle and not some other component of the entire system (disks, switches, internet connection).

    What we are left with is either trying to push systems to their limits knowing that it will not necessarily translate into real world results, or trying to perform reasonable tests that we can use to determine if a real world pilot is worthwhile and provide us with a reasonable expectation of results. This particular test has 2 characteristics that have been far too rare in all the comments I have seen on the topic across multiple boards.

    1. It is repeatable. Anyone can setup a similar situation and perform the test.
    2. It provides quantifiable results rather than "my brother's girlfriend's cousin's former roomate said...."

    While an iperf test may not be the best "real world" simulation, in this case it does provide a true apples to apples comparison between the two NICs where everything other than the NIC is held constant. The results indicate that there is a difference in the throughput of the two NICs, however the difference is relatively small. Might QoS make a difference? Yes. Might using Untangle as the gateway make a difference? Yes, but similar testing using Untangle has shown similar results. I will post those in another thread and follow up with the link.

  4. #14
    Untangle Ninja
    Join Date
    Jan 2011
    Posts
    1,327

    Default

    I'm in orange county, ca, and I've yet to see anyone who actually has a connection faster than 10mb. I'm not saying they don't exist here, I'm sure plenty of people have faster connections... it's just no one I know. The vast majority of both homes and small businesses use 3mb ADSL. Some businesses have single or dual T-1s, or a wireless equivalent.

  5. #15
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,554

    Default

    Quote Originally Posted by johnsonx42 View Post
    I'm in orange county, ca, and I've yet to see anyone who actually has a connection faster than 10mb. I'm not saying they don't exist here, I'm sure plenty of people have faster connections... it's just no one I know. The vast majority of both homes and small businesses use 3mb ADSL. Some businesses have single or dual T-1s, or a wireless equivalent.
    If that is the case then realtek is more than enough. I'm using Intel in everything that Untangle Appliances sells because I don't know what will be used. FiOS is open to a good portion of the country, DOCSIS 3.0 is spreading like fire, and I require a system to be able to survive at least three years of service. So while you may be able to get away with the realtek now, can you say the same in three years? You can't just replace NICs in the NG series.

    Also, iperf is a good indicator of base interface performance yes. That said, I've got iperf results from many RTL8111 based interfaces that all stopped at 20mb of transfer despite being synced to a 1gb switch. This was before the UVM was engaged, just Untangle with the VM stopped and running iperf through a NATless UT router or bridge.

    So if you're seeing those kinds of speed on that level, you've done something differently than I did.
    Last edited by sky-knight; 05-17-2011 at 12:46 PM.
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  6. #16
    Untangle Ninja YeOldeStonecat's Avatar
    Join Date
    Aug 2007
    Posts
    1,565

    Default

    Quote Originally Posted by johnsonx42 View Post
    I've never heard of anyone around here with such a fast connection. For whatever reasons, genuinely high-speed internet is just very rare around here.
    Docsis 3 is all over the place by me, we have clients on Comcast 50/10 and 100/10 pipes. And over in New York..similar with TWarner. We have clients on Verizons fiber..but none on the fast 160 pipe yet. Hopefully soon!

  7. #17
    Untangle Ninja YeOldeStonecat's Avatar
    Join Date
    Aug 2007
    Posts
    1,565

    Default

    Quote Originally Posted by ixeous View Post
    This is an urealistic testing standard. It is excessively difficult to truely simulate 100, 200, or 500 users .
    Unrealistic testing...yes. I've been involved in router testing since the first boxes of broadband boxes shipped out of Linksys back in the version 1 firmware days with their incredibly buggy PPPoE support and lack of MTU adjustments. I can write volumes of boring stuff on it.

    However unrealistic the testing is, those facts are realistic as far as every day use. Lets look at the facts, shall we?
    *Untangle is aimed at business networks. SMB to enterprise..and growing larger.
    *Business networks are filled with lots of different types of traffic, as mentioned above
    *Business networks are often filled with larger amounts of users, thus higher concurrent sessions.

    Routers which may fly in simple performance tests might tend to bog down under more real world scenarios. This...I can say from many years of experience in doing SMB consulting. I can also say, from installing and using Untangle on lots of different hardware since back in the version 5 days (and perhaps a hair earlier)...skimping out on NICs will get you a box of grief. I can say from experience...put in an el cheapo NIC in a business production environment....and you'll suffer the consequences. I've done an incredible amount of Untangle installations on many many many different hardware combinations, and put them into use. I've seen what works, and I've seen what doesn't work. I've had car trunk loads of different NICs go through those boxes. I've seen other experienced users here say the same thing.

    Hey, you want to embrace the Cyrix/AMD and Via/SIS bargain hardware world..go for it, but when us peeps that have experience from being in the trenches for a long time state what works and what doesn't work...may wanna save yourself the hassle. Nobody here is claiming that ""my brother's girlfriend's cousin's former roomate said...."...because we seen it ourselves....been there, done that, brought home the tee shirt.
    Last edited by YeOldeStonecat; 05-17-2011 at 05:19 PM.

  8. #18
    Master Untangler
    Join Date
    Jul 2010
    Posts
    929

    Default

    Quote Originally Posted by YeOldeStonecat View Post
    Unrealistic testing...yes. I've been involved in router testing since the first boxes of broadband boxes shipped out of Linksys back in the version 1 firmware days with their incredibly buggy PPPoE support and lack of MTU adjustments. I can write volumes of boring stuff on it.

    However unrealistic the testing is, those facts are realistic as far as every day use. Lets look at the facts, shall we?
    *Untangle is aimed at business networks. SMB to enterprise..and growing larger.
    *Business networks are filled with lots of different types of traffic, as mentioned above
    *Business networks are often filled with larger amounts of users, thus higher concurrent sessions.

    Routers which may fly in simple performance tests might tend to bog down under more real world scenarios. This...I can say from many years of experience in doing SMB consulting. I can also say, from installing and using Untangle on lots of different hardware since back in the version 5 days (and perhaps a hair earlier)...skimping out on NICs will get you a box of grief. I can say from experience...put in an el cheapo NIC in a business production environment....and you'll suffer the consequences. I've done an incredible amount of Untangle installations on many many many different hardware combinations, and put them into use. I've seen what works, and I've seen what doesn't work. I've had car trunk loads of different NICs go through those boxes. I've seen other experienced users here say the same thing.

    Hey, you want to embrace the Cyrix/AMD and Via/SIS bargain hardware world..go for it, but when us peeps that have experience from being in the trenches for a long time state what works and what doesn't work...may wanna save yourself the hassle. Nobody here is claiming that ""my brother's girlfriend's cousin's former roomate said...."...because we seen it ourselves....been there, done that, brought home the tee shirt.
    Wasn't it you that told me to us intel nic's and well i listened and have NOT had any problems at all with any of my untangle units.

  9. #19
    Untanglit ixeous's Avatar
    Join Date
    Sep 2007
    Posts
    24

    Default Silly me

    First: The link from earlier. http://forums.untangle.com/hardware/...-part-1-a.html

    Second:

    Quote Originally Posted by YeOldeStonecat View Post
    Unrealistic testing...yes. I've been involved in router testing since the first boxes of broadband boxes shipped out of Linksys back in the version 1 firmware days with their incredibly buggy PPPoE support and lack of MTU adjustments. I can write volumes of boring stuff on it.

    However unrealistic the testing is, those facts are realistic as far as every day use. Lets look at the facts, shall we?
    *Untangle is aimed at business networks. SMB to enterprise..and growing larger.
    *Business networks are filled with lots of different types of traffic, as mentioned above
    *Business networks are often filled with larger amounts of users, thus higher concurrent sessions.

    Routers which may fly in simple performance tests might tend to bog down under more real world scenarios. This...I can say from many years of experience in doing SMB consulting. I can also say, from installing and using Untangle on lots of different hardware since back in the version 5 days (and perhaps a hair earlier)...skimping out on NICs will get you a box of grief. I can say from experience...put in an el cheapo NIC in a business production environment....and you'll suffer the consequences. I've done an incredible amount of Untangle installations on many many many different hardware combinations, and put them into use. I've seen what works, and I've seen what doesn't work. I've had car trunk loads of different NICs go through those boxes. I've seen other experienced users here say the same thing.

    Hey, you want to embrace the Cyrix/AMD and Via/SIS bargain hardware world..go for it, but when us peeps that have experience from being in the trenches for a long time state what works and what doesn't work...may wanna save yourself the hassle. Nobody here is claiming that ""my brother's girlfriend's cousin's former roomate said...."...because we seen it ourselves....been there, done that, brought home the tee shirt.
    I apologize. I should have realized that a quantitative comparison is no match for "car trunk loads of different NICs". Also, it's not as if any "peeps that have experience" have ever made claims that were, over time, shown to be no longer accurate. I personally don't remember a significant number of people saying that x86 servers can't compete with big iron for mission critical tasks. What does the server market look like today? Furthermore, any work I have done with Untangle since version 5 certainly does not count as "experience", nor does over a decade working in IT for organizations that tend to be somewhat larger than a typical SMB.

    Additionally, I apologize for trying to share some specific information about performance which in no way implied anything about reliability. It was foolish of me to think that anyone may be interested in knowing what types of hardware could fulfill their performance needs.

  10. #20
    Untangle Ninja hescominsoon's Avatar
    Join Date
    Sep 2007
    Posts
    1,704

    Default

    you're being a drama queen. Most of us here have real time...years of experience with real hardware nics(intel/ broadcom) and soft nics(realtek, linksys, dlink, marvel)...they simply don't hold up with UT or any other business environment that's driving any real traffic(10 megabits or higher). In lesser environments yeah they'll work..for a while. I've had issues with soft nics in my Astaro installations and Astaro isn't nearly as memory heavy as the UVm in UT is. I go hardware nics or go home now. If somebody wants to use softnics..i'll warn them..then gladly charge them full price for the consulation fee AND the replacement hardware nic once the softnic barfs.

Page 2 of 8 FirstFirst 1234 ... LastLast

Posting Permissions

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

SEO by vBSEO 3.6.0 PL2