Announcement

Collapse

Fandango at Home Forum Guidelines

Fandango at Home Forum Guidelines

The Fandango at Home Forums are designed to help viewers get the most out of their Fandango at Home experience. Here, Fandango at Home customers may post information, questions, ideas, etc. on the subject of Fandango at Home and Fandango at Home -related issues (home theater, entertainment, etc). Although the primary purpose of these forums is to help Fandango at Home customers with questions and/or problems with their Fandango at Home service, there are also off-topic areas available within the Fandango at Home Forums for users to chat with like-minded people, subject to the limitations below.

Please post all comments in English. When posting a comment in the Fandango at Home Forums, please conduct yourself in a respectful and civil manner. While we respect that you may feel strongly about an issue, please leave room for discussion.

Fandango at Home Forum Guidelines

The Fandango at Home Forums are designed to help viewers get the most out of their Fandango at Home experience. Here, Fandango at Home customers may post information, questions, ideas, etc. on the subject of Fandango at Home and Fandango at Home -related issues (home theater, entertainment, etc). Although the primary purpose of these forums is to help Fandango at Home customers with questions and/or problems with their Fandango at Home service, there are also off-topic areas available within the Fandango at Home Forums for users to chat with like-minded people, subject to the limitations below.

Please post all comments in English. When posting a comment in the Fandango at Home Forums, please conduct yourself in a respectful and civil manner. While we respect that you may feel strongly about an issue, please leave room for discussion.

Fandango at Home reserves the right to refrain from posting and/or to remove user comments, including comments that contain any of the following:

1. Obscenities, defamatory language, discriminatory language, or other language not suitable for a public forum
2. Email addresses, phone numbers, links to websites, physical addresses or other forms of contact information
3. "Spam" content, references to other products, advertisements, or other offers
4. Spiteful or inflammatory comments about other users or their comments
5. Comments that may potentially violate the DMCA or any other applicable laws
6. Comments that discuss ways to manipulate Fandango at Home products/services, including, but not limited to, reverse engineering, video extraction, and file conversion.

Additionally, please keep in mind that although Fandango at Home retains the right to monitor, edit, and/or remove posts within Fandango at Home Forums, it does not necessarily review every comment. Accordingly, specific questions about Fandango at Home products and services should be directed to Fandango at Home customer service representatives.

Terms of Use - User Comments, Feedback, Reviews, Submissions

For all reviews, comments, feedback, postcards, suggestions, ideas, and other submissions disclosed, submitted or offered to Fandango at Home, on or through this Site, by e-mail or telephone, or otherwise disclosed, submitted or offered in connection you use of this Site (collectively, the "Comments") you grant Fandango at Home a royalty-free, irrevocable, transferable right and license to use the Comments however Fandango at Home desires, including, without limitation, to copy, modify, delete in its entirety, adapt, publish, translate, create derivative works from and/or sell and /or distribute such Comments and/or incorporate such Comments into any form, medium or technology throughout the world.
Fandango at Home will be entitled to use, reproduce, disclose, modify, adapt, create derivative works from, publish, display and distribute any Comments you submit for any purpose whatsoever, without restriction and without compensating you in any way. Fandango at Home is and shall be under no obligation (1) to maintain any Comments in confidence; (2) to pay to users any compensation for any Comments; or (3) to respond to any user Comments. You agree that any Comments submitted by you to the Site will not violate the terms in this Terms of Use or any right of any third party, including without limitation, copyright, trademark, privacy or other personal or proprietary right(s), and will not cause injury to any person or entity. You further agree that no Comments submitted by you to this Site will be or contain libelous or otherwise unlawful, threatening, abusive or obscene material, or contain software viruses, political campaigning, commercial solicitation, chain letters, mass mailings or any form of "spam."

You grant Fandango at Home the right to use the name that you submit in connection with any Comments. You agree not to use a false email address, impersonate any person or entity, otherwise mislead as to the origin of any Comments you submit. You are, and shall remain, solely responsible for the content of any Comments you make and you agree to indemnify Fandango at Home for all claims resulting from any Comments you submit. Fandango at Home takes no responsibility and assumes no liability for any Comments submitted by you or any third-party reserves the right to refrain from posting and/or to remove user comments, including comments that contain any of the following:

1. Obscenities, defamatory language, discriminatory language, or other language not suitable for a public forum
2. Email addresses, phone numbers, links to websites, physical addresses or other forms of contact information
3. "Spam" content, references to other products, advertisements, or other offers
4. Spiteful or inflammatory comments about other users or their comments
5. Comments that may potentially violate the DMCA or any other applicable laws
6. Comments that discuss ways to manipulate Fandango at Home products/services, including, but not limited to, reverse engineering, video extraction, and file conversion.

Additionally, please keep in mind that although Fandango at Home retains the right to monitor, edit, and/or remove posts within Fandango at Home Forums, it does not necessarily review every comment. Accordingly, specific questions about Fandango at Home products and services should be directed to Fandango at Home customer service representatives.

Terms of Use - User Comments, Feedback, Reviews, Submissions

For all reviews, comments, feedback, postcards, suggestions, ideas, and other submissions disclosed, submitted or offered to Fandango at Home, on or through this Site, by e-mail or telephone, or otherwise disclosed, submitted or offered in connection you use of this Site (collectively, the "Comments") you grant Fandango at Home a royalty-free, irrevocable, transferable right and license to use the Comments however Fandango at Home desires, including, without limitation, to copy, modify, delete in its entirety, adapt, publish, translate, create derivative works from and/or sell and /or distribute such Comments and/or incorporate such Comments into any form, medium or technology throughout the world.
Fandango at Home will be entitled to use, reproduce, disclose, modify, adapt, create derivative works from, publish, display and distribute any Comments you submit for any purpose whatsoever, without restriction and without compensating you in any way. Fandango at Home is and shall be under no obligation (1) to maintain any Comments in confidence; (2) to pay to users any compensation for any Comments; or (3) to respond to any user Comments. You agree that any Comments submitted by you to the Site will not violate the terms in this Terms of Use or any right of any third party, including without limitation, copyright, trademark, privacy or other personal or proprietary right(s), and will not cause injury to any person or entity. You further agree that no Comments submitted by you to this Site will be or contain libelous or otherwise unlawful, threatening, abusive or obscene material, or contain software viruses, political campaigning, commercial solicitation, chain letters, mass mailings or any form of "spam."

You grant Fandango at Home the right to use the name that you submit in connection with any Comments. You agree not to use a false email address, impersonate any person or entity, otherwise mislead as to the origin of any Comments you submit. You are, and shall remain, solely responsible for the content of any Comments you make and you agree to indemnify Fandango at Home for all claims resulting from any Comments you submit. Fandango at Home takes no responsibility and assumes no liability for any Comments submitted by you or any third-party.
See more
See less

IGMP 10/100/1000 Ethernet Switch

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    #16
    Re: IGMP 10/100/1000 Ethernet Switch

    Originally posted by rstone View Post
    The maximum theoretical throughput is 125MB/s (1000 Mbps / 8 = 125 MB/s) with Gigabit speed over a single ethernet connection. The reality is that your probably more likely to be limited by the limitations of your hard drive, CPU, PCI bus, etc, than your network. You can always try teaming with a switch that has 802.3ad (Link Aggregation) to improve speeds over multiple ethernet cables. Regardless 100 MB/sec is actually good, so I wouldn't worry about it.

    I'm getting 100 Mb/sec not 100 MB/sec - SMALL b. That's the problem...I'm now running with a Netgear ProSafe 8 port switch. And my CPUs have plenty of juice. Maybe I'm not going to get a full 100 MB/sec but I should get more than what I am getting.

    Maybe I need to start a new thread on this...

    Comment


      #17
      Re: IGMP 10/100/1000 Ethernet Switch

      Originally posted by rstone View Post
      Is that over a ethernet network? Just curious, you running Fiber? 10GigE network??

      My guess is what your seeing is MBit per second, not MB/sec so 650 MBit / 8 = 81.25 MB/sec
      No fiber yet....though about it 10 years ago but it was cost prohibitative. But yes I am using bits per second not bytes. When networking I always use small "b"...

      Comment


        #18
        Re: IGMP 10/100/1000 Ethernet Switch

        Originally posted by NA9D View Post
        I'm getting 100 Mb/sec not 100 MB/sec - SMALL b. That's the problem...I'm now running with a Netgear ProSafe 8 port switch. And my CPUs have plenty of juice. Maybe I'm not going to get a full 100 MB/sec but I should get more than what I am getting.

        Maybe I need to start a new thread on this...
        If you are only getting 12.5 MB/sec (100mbps) that is pretty slow for gigabit. Are you running gigabit with jumbo frames? that should push you into the 20+ MB/sec range, if you using 100 MB/sec network than the 12.5 MB/sec is to be expected if you figure that 100 / 8 = 10 MB/sec, but with gigabit you should be seeing alot higher speeds.

        Comment


          #19
          Re: IGMP 10/100/1000 Ethernet Switch

          Originally posted by NA9D View Post
          I'm getting 100 Mb/sec not 100 MB/sec - SMALL b. That's the problem...I'm now running with a Netgear ProSafe 8 port switch. And my CPUs have plenty of juice. Maybe I'm not going to get a full 100 MB/sec but I should get more than what I am getting.

          Maybe I need to start a new thread on this...
          Could you have a damaged cable? If one pair is damaged it will fall back to 100mbs since it needs all four pairs to run at gigabit.

          It could be something else too. I lost gigabit capability in one of my laptops after I had a static discharge when I picked it up. So then it would only connect at 100mbs. Then after the next static discharge on the laptop when I touched it, a few weeks later, I lost all wired connectivity. I need to send it back for warranty repair since even with wireless N(5Ghz) at over 100mbs throughput, it takes too long to transfer 100GB of data.


          The reason I switched to 1000mbs in 2001 from 100mbs was for transfers of my HD recordings. it would take too long to transfer the 50GB of HD recordings(I recorded at another location and would transfer the data from a removable hard drive to storage on my network) I had each week on a 100mbs network. Switching o 1000mbs sped things up tremendously.

          Also it doesn't take much of a PC to run at gigabit. Even in 2001 when I was using a P3 CPU at 667mhz, i had no problem getting several hundred megabits throughput.

          Comment


            #20
            Re: IGMP 10/100/1000 Ethernet Switch

            Originally posted by RonV View Post
            aaronwt, you may want to read this paper on how IGMP & IP Multicasting works. Like I said before, if a switch doesn't support the protocol natively then it broadcasts the stream out all ports..aka switchkill.

            http://www.juniper.net/solutions/lit...ers/200188.pdf

            Thanks!

            Comment


              #21
              Re: IGMP 10/100/1000 Ethernet Switch

              Originally posted by RonV View Post
              No fiber yet....though about it 10 years ago but it was cost prohibitative. But yes I am using bits per second not bytes. When networking I always use small "b"...
              Got ya. I'll keep that in mind when you talk about performance numbers. Being that I use gigabit and high performance hardware which exceeds 100 MB/sec, and that most hardware (ie. SATA, SAS, PCI, PCI-E) is reported in bytes, and even VISTA reports numbers in MB/sec, I tend to use bytes when talking about performance. I think the only time I ever refer to bits is when talking about internet speeds.

              Comment


                #22
                Re: IGMP 10/100/1000 Ethernet Switch

                All my equipment states that the connection is at 1 Gigabit/second. It all links up that way - I just don't get the transfer rate...

                Comment


                  #23
                  Re: IGMP 10/100/1000 Ethernet Switch

                  Originally posted by NA9D View Post
                  All my equipment states that the connection is at 1 Gigabit/second. It all links up that way - I just don't get the transfer rate...
                  Could be that you don't have jumbo frames enabled, on your NIC and switch. In VISTA goto Network / Network and Sharing Center / Manage Network Connections / right click on your connection and select properties / click on configure / find your advance settings / you should see something that says something like Jumbo Packets, and then when selected you will see numbers, select the highest one, usually somewhere between 7k and 9k. Then make sure your switch support Jumbo Frames, check with manufacture, if it doesn't you will need to swap your switch out, if it does and it is a dumb switch, you don't need to set anything, if it has a web interface, make sure jumbo frames is enabled (some are off by default).

                  Also, try these things to help improve performance in VISTA, if you haven't already.
                  How to use the throttling mechanism to control network performance in Windows Vista

                  To configure the throttling rate, change the value of the NetworkThrottlingIndex registry entry under the following registry subkey:

                  HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Multimedia\SystemProfile\
                  Name : NetworkThrottlingIndex
                  Value type : DWORD
                  Value data : From integer 1 through integer 70 (Decimal) (Decimal)

                  By default, the value for the NetworkThrottlingIndex registry entry is set to 10. When the NetworkThrottlingIndex registry entry does not exist, the behavior resembles the default behavior. Network throttling can be completely turned off by setting the value to FFFFFFFF (hexadecimal). You must restart the computer after you make a change to the value of the NetworkThrottlingIndex registry entry.

                  Also under VISTA, if you goto Control Panel / Add programs and features / Turn windows features on and off / and turn off Remote Differential Compression.

                  http://weblog.empoweris.com/2007/09/...l-compression/

                  What is RDC you may ask? Below is a quote and explanation from Microsoft.
                  RDC is a ?diff-over-the wire? client-server protocol that can be used to efficiently update files over a limited-bandwidth network. RDC detects insertions, removals, and re-arrangements of data in files, enabling DFS Replication to replicate only the changed file blocks when files are updated.
                  In English that means Remote Differential Compression measures the changes in files over a network to transfer them with minimal bandwidth rather than transferring an entire file that has previously been moved. By constantly checking for file changes, this service can hinder system performance.

                  So does this constant searching for file updates slow your system down? It can and sometimes will. If you are looking to try to do some tuning and would like to boost your system, you should try turning off RDC. Unless your system is replicating DFS from a network, your should be good to go.

                  Comment


                    #24
                    Re: IGMP 10/100/1000 Ethernet Switch

                    Also, if none of that works, you can tweak your NIC transmit and receive buffers, and turn off/on "Offload TCP" and "Offload Checksum" to increase performance, just make sure you write your settings down ahead of time, in case it makes it worse.

                    Another thing you can do is turn off your security and anti-virus programs to see if they are bottlenecking your network.

                    Lastly, if all else fails, I'd suggest you swap out your hard drives for Western Digital Raptor drives, these drives are extremely fast, and if you pick up two and set them up in RAID 0, 1 you should see a remarkable improvement.

                    Comment


                      #25
                      Re: IGMP 10/100/1000 Ethernet Switch

                      Even without Jumbo frames you should have no problem getting several hundred mbs throughput. I don't use jumbo frames on my network.

                      Comment


                        #26
                        Re: IGMP 10/100/1000 Ethernet Switch

                        Originally posted by aaronwt View Post
                        Even without Jumbo frames you should have no problem getting several hundred mbs throughput. I don't use jumbo frames on my network.
                        If you have gigabit and your not using Jumbo Frames, your missing out...the use of Jumbo Frames can deliver a 50% increase in throughput with a simultaneous 50% decrease in CPU utilization.

                        If your concerned about error checking at higher bitrates, then just keep in mind that with today's current Ethernet that they are more than capable of detecting bit errors at frame sizes up to 12,000 bytes. So even the maximum used of 9,000 bytes by most NICs and switches, it well under the limit of what Ethernet can handle.

                        Personally, I do everything I can to increase my throughput and I easily transfer files on my network that exceeds 102 MB/sec to my PCs and 125 MB/sec from my PCs, consistently and sustainably. That is 816 / 1000 bits for you other folks.

                        This is a good read, if your interested in WHY you should use Jumbo Frames.

                        http://www.smallnetbuilder.com/conte.../30201/54/1/1/

                        Comment


                          #27
                          Re: IGMP 10/100/1000 Ethernet Switch

                          Thanks for the advise. One, I run Mac primarily and on my PCs I don't run Vista!

                          But that's OK. Thanks for the Wiki nonetheless.

                          Anyhow, how does the use of jumbo frames work with devices that don't support them (ie: like a Tivo). Is the protocol smart enough to lower the frame size for those sort of devices?

                          Second, I enabled jumbo frames on my Buffalo NAS units and on my computers and the throughput was a nightmare. I don't know if it's a problem on Buffalo's end or what. I'll try it again and see though...

                          Comment


                            #28
                            Re: IGMP 10/100/1000 Ethernet Switch

                            Originally posted by rstone View Post
                            If you have gigabit and your not using Jumbo Frames, your missing out...the use of Jumbo Frames can deliver a 50% increase in throughput with a simultaneous 50% decrease in CPU utilization.

                            If your concerned about error checking at higher bitrates, then just keep in mind that with today's current Ethernet that they are more than capable of detecting bit errors at frame sizes up to 12,000 bytes. So even the maximum used of 9,000 bytes by most NICs and switches, it well under the limit of what Ethernet can handle.

                            Personally, I do everything I can to increase my throughput and I easily transfer files on my network that exceeds 102 MB/sec to my PCs and 125 MB/sec from my PCs, consistently and sustainably. That is 816 / 1000 bits for you other folks.

                            This is a good read, if your interested in WHY you should use Jumbo Frames.

                            http://www.smallnetbuilder.com/conte.../30201/54/1/1/
                            I already get up to 850mbs throughput and CPU utilzation is negligible on my dual core and quad core machines.
                            Although I have tried jumbo frames in the past and it worked fine for me with large file transfers while small file transfers were slower(I'm talking like a gigabyte of very small files). In the end I kept it at the default since that gave me the best overall performance when I tested all my options on my network.

                            edit: that CPU utilization test is with dual 300mhz Sun servers. My main PCs are between 2Ghz dual core up to 3.3Ghz quad core with one 2.6ghz single core for my Windows Home Server and a 1.5Ghz single core on an old laptop.

                            Comment


                              #29
                              Re: IGMP 10/100/1000 Ethernet Switch

                              Originally posted by NA9D View Post
                              Anyhow, how does the use of jumbo frames work with devices that don't support them (ie: like a Tivo). Is the protocol smart enough to lower the frame size for those sort of devices?

                              Second, I enabled jumbo frames on my Buffalo NAS units and on my computers and the throughput was a nightmare. I don't know if it's a problem on Buffalo's end or what. I'll try it again and see though...
                              Well the general consensus is that you really should isolate the two onto two different switches, one with Jumbo Frames and one without. Personally, I have found it depends on the switch. My switch handles both fine. I would say, if you want to try out Jumbo Frames, then first try it with everything on one switch, if it cause problems or slows throughput on your non-jumbo frames machines, then you probably need separate switches. That might also explain the problem with your Buffalo NAS as well, but I know on my NAS with/without Jumbo Frames can easily make a 20+ MB/sec difference.

                              As for you being a mac user, those tweaking your NIC transmit and receive buffers, and turn off/on "Offload TCP" and "Offload Checksum" to increase performance still applies. And if you have a MAC that you can swap out HDDs, then putting faster drives should also apply as well.

                              Comment


                                #30
                                Re: IGMP 10/100/1000 Ethernet Switch

                                Originally posted by aaronwt View Post
                                I already get up to 850mbs throughput and CPU utilzation is negligible on my dual core and quad core machines.
                                Although I have tried jumbo frames in the past and it worked fine for me with large file transfers while small file transfers were slower(I'm talking like a gigabyte of very small files). In the end I kept it at the default since that gave me the best overall performance when I tested all my options on my network.

                                edit: that CPU utilization test is with dual 300mhz Sun servers. My main PCs are between 2Ghz dual core up to 3.3Ghz quad core with one 2.6ghz single core for my Windows Home Server and a 1.5Ghz single core on an old laptop.
                                Well the way I look at it is, ever mbps counts. Even if you only get 1% increase, that's 1% you didn't have before. Regardless, 850 mbps is nothing to complain about.

                                You don't need a dual 300mhz Sun server to see a 50% increase in CPU utilization tests, in fact when I setup new test machine, just recently to do some network performance testing, I tweaked the NICs on it, so that I easily saw a 50% increase in CPU utilization, which also increased network transfer speeds.

                                Personally, I don't transfer much, if anything between individual PCs, that why I have an NAS. Only my server gets some usage from other PCs, and that is mostly internet traffic for the server Website, DNS, Mail, etc, but even so I really don't transfer file from/to the server directly from the other PCs on the network, except maybe on a rare occasion.

                                Comment

                                Working...
                                X