Why is FileZilla SFTP file transfer max capped at 1.3MiB/sec instead of saturating available bandwidth? rsync and WinSCP are even slowerSpeed up SFTP uploads on high latency network?How to use openssh sftp command with a RSA/DSA key specified from the command lineTransfer large files over the internet from a Linux VPS to Windows machineWinSCP - SCP is working and SFTP is not workingOpenSSH anything like 'internal-sftp' but for SCP?Is it possible to force SFTP rather than allow SCP fallback ?Two factor authentication for SSHCopying over SSH via Putty tools is slower than via WinSCPReplace scp with sftpHow to leave connection to SFTP open, while other bash commands runLoad Testing Scenario for upload/download file to/from sftp server using winscp client on Azure

Why wasn't the Night King naked in S08E03?

What are the advantages of luxury car brands like Acura/Lexus over their sibling non-luxury brands Honda/Toyota?

How do inspiraling black holes get closer?

Understanding trademark infringements in a world where many dictionary words are trademarks?

How should I tell my manager I'm not paying for an optional after work event I'm not going to?

Do I add my skill check modifier to the roll of 15 granted by Glibness?

Why is "breaking the mould" positively connoted?

What exactly are the `size issues' preventing formation of presheaves being a left adjoint to some forgetful functor?

Can there be a single technologically advanced nation, in a continent full of non-technologically advanced nations?

When two pilots are required for a private aircraft, is it a requirement for the PIC to be ATPL?

Decoupling cap routing on a 4 layer PCB

A factorization game

Upside-Down Pyramid Addition...REVERSED!

Why aren't nationalizations in Russia described as socialist?

Does the 7th major scale note resolve more strongly to the lower tonic (note 1) than the higher tonic (note 8)?

Would glacier 'trees' be plausible?

Should I decline this job offer that requires relocating to an area with high cost of living?

Can my company stop me from working overtime?

Can my 2 children 10 and 12 Travel to the USA on expired American Passports? They are US citizens

Can you Ready a Bard spell to release it after using Battle Magic?

60s/70s science fiction novel where a man (after years of trying) finally succeeds to make a coin levitate by sheer concentration

Find the cheapest shipping option based on item weight

How did the Venus Express detect lightning?

Why does this derived table improve performance?



Why is FileZilla SFTP file transfer max capped at 1.3MiB/sec instead of saturating available bandwidth? rsync and WinSCP are even slower


Speed up SFTP uploads on high latency network?How to use openssh sftp command with a RSA/DSA key specified from the command lineTransfer large files over the internet from a Linux VPS to Windows machineWinSCP - SCP is working and SFTP is not workingOpenSSH anything like 'internal-sftp' but for SCP?Is it possible to force SFTP rather than allow SCP fallback ?Two factor authentication for SSHCopying over SSH via Putty tools is slower than via WinSCPReplace scp with sftpHow to leave connection to SFTP open, while other bash commands runLoad Testing Scenario for upload/download file to/from sftp server using winscp client on Azure






.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty height:90px;width:728px;box-sizing:border-box;








5















I'm downloading from a server and downloads are maxing out at 1.3MiB/second with FileZilla but I can start concurrent downloads and they will download at 1.3MiB/second also. So why can't I download just one file at faster than 1.3MB/s and get closer to saturate available bandwidth (~6+MB/s)?



I know that I can use some other SFTP client that supports segmented downloads such as lftp, know of other good ones that are open source?



But I still want to know what is it that limits downloading one file to just 1.3MB/s, is it some technical limitation with TCP and buffers etc or some configuration issue? I checked and for sure there's no traffic throttling enabled at all for FileZilla.



Also I tried rsync and it was worse than FileZilla/SFTP. I also tried WinSCP and it was the slowest regardless of method SCP/SFTP. So at 1.3MB/s constant transfer FileZilla is pretty good compared to the other methods of transfer.



If someone has a good explanation of why transfers peak at 1.3MB/s I'd really like to know, and if its possible to increase this without resorting to using segmented downloading. Server is running OpenSSH 6.7p1 (Debian) client is FileZilla on Windows.



UPDATE: In response to Martin's information (see his answer below) I am adding that ping is 180ms to 190ms pretty constant between server and client that is downloading. Also cpu usage is very low, 2% to 8% max.
I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also.



So could the 180ms delay to the server be a mathematically limiting factor as Martin and Michael touched on a bit? Or could there still be something else to blame such that I can improve the throughput? If not, I'd appreciate if anyone knows any other (like lftp but runs well on Windows) open source downloader that is secure and supports segmented downloading.










share|improve this question



















  • 6





    Look up "bandwidth-delay product".

    – Michael Hampton
    May 24 '15 at 6:18











  • Hi, thanks I'll look that up, please see my reply to Martin and my updated question.

    – htfree
    May 24 '15 at 19:58











  • I switched to FTPS, that helped. also as one user said changing priority of filezilla thread to high helped a bit seems. Perhaps I will try ssh sftp mode again once I setup HPN-SSH ...

    – htfree
    Jul 23 '16 at 3:33











  • HPN-SSH + bdp is the answer. Also see fasterdata.es.net

    – Dan Pritts
    Sep 5 '17 at 20:07

















5















I'm downloading from a server and downloads are maxing out at 1.3MiB/second with FileZilla but I can start concurrent downloads and they will download at 1.3MiB/second also. So why can't I download just one file at faster than 1.3MB/s and get closer to saturate available bandwidth (~6+MB/s)?



I know that I can use some other SFTP client that supports segmented downloads such as lftp, know of other good ones that are open source?



But I still want to know what is it that limits downloading one file to just 1.3MB/s, is it some technical limitation with TCP and buffers etc or some configuration issue? I checked and for sure there's no traffic throttling enabled at all for FileZilla.



Also I tried rsync and it was worse than FileZilla/SFTP. I also tried WinSCP and it was the slowest regardless of method SCP/SFTP. So at 1.3MB/s constant transfer FileZilla is pretty good compared to the other methods of transfer.



If someone has a good explanation of why transfers peak at 1.3MB/s I'd really like to know, and if its possible to increase this without resorting to using segmented downloading. Server is running OpenSSH 6.7p1 (Debian) client is FileZilla on Windows.



UPDATE: In response to Martin's information (see his answer below) I am adding that ping is 180ms to 190ms pretty constant between server and client that is downloading. Also cpu usage is very low, 2% to 8% max.
I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also.



So could the 180ms delay to the server be a mathematically limiting factor as Martin and Michael touched on a bit? Or could there still be something else to blame such that I can improve the throughput? If not, I'd appreciate if anyone knows any other (like lftp but runs well on Windows) open source downloader that is secure and supports segmented downloading.










share|improve this question



















  • 6





    Look up "bandwidth-delay product".

    – Michael Hampton
    May 24 '15 at 6:18











  • Hi, thanks I'll look that up, please see my reply to Martin and my updated question.

    – htfree
    May 24 '15 at 19:58











  • I switched to FTPS, that helped. also as one user said changing priority of filezilla thread to high helped a bit seems. Perhaps I will try ssh sftp mode again once I setup HPN-SSH ...

    – htfree
    Jul 23 '16 at 3:33











  • HPN-SSH + bdp is the answer. Also see fasterdata.es.net

    – Dan Pritts
    Sep 5 '17 at 20:07













5












5








5


4






I'm downloading from a server and downloads are maxing out at 1.3MiB/second with FileZilla but I can start concurrent downloads and they will download at 1.3MiB/second also. So why can't I download just one file at faster than 1.3MB/s and get closer to saturate available bandwidth (~6+MB/s)?



I know that I can use some other SFTP client that supports segmented downloads such as lftp, know of other good ones that are open source?



But I still want to know what is it that limits downloading one file to just 1.3MB/s, is it some technical limitation with TCP and buffers etc or some configuration issue? I checked and for sure there's no traffic throttling enabled at all for FileZilla.



Also I tried rsync and it was worse than FileZilla/SFTP. I also tried WinSCP and it was the slowest regardless of method SCP/SFTP. So at 1.3MB/s constant transfer FileZilla is pretty good compared to the other methods of transfer.



If someone has a good explanation of why transfers peak at 1.3MB/s I'd really like to know, and if its possible to increase this without resorting to using segmented downloading. Server is running OpenSSH 6.7p1 (Debian) client is FileZilla on Windows.



UPDATE: In response to Martin's information (see his answer below) I am adding that ping is 180ms to 190ms pretty constant between server and client that is downloading. Also cpu usage is very low, 2% to 8% max.
I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also.



So could the 180ms delay to the server be a mathematically limiting factor as Martin and Michael touched on a bit? Or could there still be something else to blame such that I can improve the throughput? If not, I'd appreciate if anyone knows any other (like lftp but runs well on Windows) open source downloader that is secure and supports segmented downloading.










share|improve this question
















I'm downloading from a server and downloads are maxing out at 1.3MiB/second with FileZilla but I can start concurrent downloads and they will download at 1.3MiB/second also. So why can't I download just one file at faster than 1.3MB/s and get closer to saturate available bandwidth (~6+MB/s)?



I know that I can use some other SFTP client that supports segmented downloads such as lftp, know of other good ones that are open source?



But I still want to know what is it that limits downloading one file to just 1.3MB/s, is it some technical limitation with TCP and buffers etc or some configuration issue? I checked and for sure there's no traffic throttling enabled at all for FileZilla.



Also I tried rsync and it was worse than FileZilla/SFTP. I also tried WinSCP and it was the slowest regardless of method SCP/SFTP. So at 1.3MB/s constant transfer FileZilla is pretty good compared to the other methods of transfer.



If someone has a good explanation of why transfers peak at 1.3MB/s I'd really like to know, and if its possible to increase this without resorting to using segmented downloading. Server is running OpenSSH 6.7p1 (Debian) client is FileZilla on Windows.



UPDATE: In response to Martin's information (see his answer below) I am adding that ping is 180ms to 190ms pretty constant between server and client that is downloading. Also cpu usage is very low, 2% to 8% max.
I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also.



So could the 180ms delay to the server be a mathematically limiting factor as Martin and Michael touched on a bit? Or could there still be something else to blame such that I can improve the throughput? If not, I'd appreciate if anyone knows any other (like lftp but runs well on Windows) open source downloader that is secure and supports segmented downloading.







ssh rsync sftp scp winscp






share|improve this question















share|improve this question













share|improve this question




share|improve this question








edited May 24 '15 at 19:58







htfree

















asked May 24 '15 at 6:12









htfreehtfree

1363616




1363616







  • 6





    Look up "bandwidth-delay product".

    – Michael Hampton
    May 24 '15 at 6:18











  • Hi, thanks I'll look that up, please see my reply to Martin and my updated question.

    – htfree
    May 24 '15 at 19:58











  • I switched to FTPS, that helped. also as one user said changing priority of filezilla thread to high helped a bit seems. Perhaps I will try ssh sftp mode again once I setup HPN-SSH ...

    – htfree
    Jul 23 '16 at 3:33











  • HPN-SSH + bdp is the answer. Also see fasterdata.es.net

    – Dan Pritts
    Sep 5 '17 at 20:07












  • 6





    Look up "bandwidth-delay product".

    – Michael Hampton
    May 24 '15 at 6:18











  • Hi, thanks I'll look that up, please see my reply to Martin and my updated question.

    – htfree
    May 24 '15 at 19:58











  • I switched to FTPS, that helped. also as one user said changing priority of filezilla thread to high helped a bit seems. Perhaps I will try ssh sftp mode again once I setup HPN-SSH ...

    – htfree
    Jul 23 '16 at 3:33











  • HPN-SSH + bdp is the answer. Also see fasterdata.es.net

    – Dan Pritts
    Sep 5 '17 at 20:07







6




6





Look up "bandwidth-delay product".

– Michael Hampton
May 24 '15 at 6:18





Look up "bandwidth-delay product".

– Michael Hampton
May 24 '15 at 6:18













Hi, thanks I'll look that up, please see my reply to Martin and my updated question.

– htfree
May 24 '15 at 19:58





Hi, thanks I'll look that up, please see my reply to Martin and my updated question.

– htfree
May 24 '15 at 19:58













I switched to FTPS, that helped. also as one user said changing priority of filezilla thread to high helped a bit seems. Perhaps I will try ssh sftp mode again once I setup HPN-SSH ...

– htfree
Jul 23 '16 at 3:33





I switched to FTPS, that helped. also as one user said changing priority of filezilla thread to high helped a bit seems. Perhaps I will try ssh sftp mode again once I setup HPN-SSH ...

– htfree
Jul 23 '16 at 3:33













HPN-SSH + bdp is the answer. Also see fasterdata.es.net

– Dan Pritts
Sep 5 '17 at 20:07





HPN-SSH + bdp is the answer. Also see fasterdata.es.net

– Dan Pritts
Sep 5 '17 at 20:07










3 Answers
3






active

oldest

votes


















11














There are three common factors that affect a transfer speed:



  • Bandwidth - An obvious factor that's apparently not your trouble.



  • Network delay/latency - The SFTP is packet oriented-protocol. When downloading, the SFTP client sends a "read" request to the SFTP server, waits for a response, appends returned data to a local file; and repeats, until the end of the file.



    Even if your connection is fast, if the server is far away (or slow), it takes a time for the data to arrive back. If the client spends this time uselessly waiting, your transfer speed will be low.



    Most SFTP clients (including FileZilla and WinSCP) overcome the problem by both requesting a large chunk of the file in each single "read" request and by sending (queuing) multiple "read" requests without waiting for a response to previous. For example WinSCP can request up to 32 chunks for 32 KB each at once, totaling 1 MB (these are defaults). But if there's a big discrepancy between the bandwidth and the network delay, even that 1 MB can be too small to saturate the bandwidth.



    An underlying TCP protocol can suffer a similar problem. So it's not only how the actual SFTP client is efficient, but also how an underlying TCP layer is efficient.



    See also Bandwidth-delay product on Wikipedia.



    I do not think this is your trouble either, at least if you have used the latest version of WinSCP for the tests. There have been some improvements in the recent releases, which allow WinSCP to utilize high-latency connections as efficiently as FileZilla.




  • CPU - The SFTP, being encrypted, is CPU intensive. If you have a relatively slow CPU, comparing to a large bandwidth, the transfer can be capped by your CPU being unable to encrypt (or decrypt in case of the download) the data as fast as your network is capable of transferring them.



    Common SFTP clients cannot distribute the encryption/decryption among CPU cores, so it's actually a capacity of a single CPU core that limits the transfer speed.



    Use the Windows Task manager to see, if one of the cores is utilized to its maximum during the transfer.







share|improve this answer




















  • 2





    1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

    – Håkan Lindqvist
    May 24 '15 at 11:08






  • 1





    Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

    – htfree
    May 24 '15 at 19:53











  • I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

    – htfree
    May 25 '15 at 21:42











  • I cannot confirm that without seeing a log file. I just tried to show you all the options.

    – Martin Prikryl
    May 26 '15 at 5:15











  • What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

    – htfree
    May 27 '15 at 2:01


















2














I had this issue as well.



I Used task manager to set the priority to high.



Now I get up to 5 MiB/s






share|improve this answer























  • I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

    – htfree
    Jul 23 '16 at 3:32


















0














I recently tried on same exact network with windows 10 and perhaps newer version of filezilla and I got up to 7MB/second transfers from the same server! I then tested with RSYNC inside a virtual machine and got 7MB/second also. I am "pretty sure" now that the problem lies with the COMODO firewall I have installed on this Windows 7 system.



Apparently even if you "disable" it, all it does is not enforce rules but it slows down network stack. I have installed/replicated this windows 7 system inside a virtual machine also and I will try to completely "remove" Comodo cis premium (antivirus+firewall) and confirm here. I should also mention on this machine I also noticed erratic intermittent latency pings to some systems on my network where all other systems between that were stable <1ms. So bandwidth delay product info is very good but in my case I was able to get filezilla and rsync both at 7MB/s (which basically saturates my available bandwidth) on another install, same network local and remote.






share|improve this answer























    Your Answer








    StackExchange.ready(function()
    var channelOptions =
    tags: "".split(" "),
    id: "2"
    ;
    initTagRenderer("".split(" "), "".split(" "), channelOptions);

    StackExchange.using("externalEditor", function()
    // Have to fire editor after snippets, if snippets enabled
    if (StackExchange.settings.snippets.snippetsEnabled)
    StackExchange.using("snippets", function()
    createEditor();
    );

    else
    createEditor();

    );

    function createEditor()
    StackExchange.prepareEditor(
    heartbeatType: 'answer',
    autoActivateHeartbeat: false,
    convertImagesToLinks: true,
    noModals: true,
    showLowRepImageUploadWarning: true,
    reputationToPostImages: 10,
    bindNavPrevention: true,
    postfix: "",
    imageUploader:
    brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
    contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
    allowUrls: true
    ,
    onDemand: true,
    discardSelector: ".discard-answer"
    ,immediatelyShowMarkdownHelp:true
    );



    );













    draft saved

    draft discarded


















    StackExchange.ready(
    function ()
    StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fserverfault.com%2fquestions%2f694062%2fwhy-is-filezilla-sftp-file-transfer-max-capped-at-1-3mib-sec-instead-of-saturati%23new-answer', 'question_page');

    );

    Post as a guest















    Required, but never shown

























    3 Answers
    3






    active

    oldest

    votes








    3 Answers
    3






    active

    oldest

    votes









    active

    oldest

    votes






    active

    oldest

    votes









    11














    There are three common factors that affect a transfer speed:



    • Bandwidth - An obvious factor that's apparently not your trouble.



    • Network delay/latency - The SFTP is packet oriented-protocol. When downloading, the SFTP client sends a "read" request to the SFTP server, waits for a response, appends returned data to a local file; and repeats, until the end of the file.



      Even if your connection is fast, if the server is far away (or slow), it takes a time for the data to arrive back. If the client spends this time uselessly waiting, your transfer speed will be low.



      Most SFTP clients (including FileZilla and WinSCP) overcome the problem by both requesting a large chunk of the file in each single "read" request and by sending (queuing) multiple "read" requests without waiting for a response to previous. For example WinSCP can request up to 32 chunks for 32 KB each at once, totaling 1 MB (these are defaults). But if there's a big discrepancy between the bandwidth and the network delay, even that 1 MB can be too small to saturate the bandwidth.



      An underlying TCP protocol can suffer a similar problem. So it's not only how the actual SFTP client is efficient, but also how an underlying TCP layer is efficient.



      See also Bandwidth-delay product on Wikipedia.



      I do not think this is your trouble either, at least if you have used the latest version of WinSCP for the tests. There have been some improvements in the recent releases, which allow WinSCP to utilize high-latency connections as efficiently as FileZilla.




    • CPU - The SFTP, being encrypted, is CPU intensive. If you have a relatively slow CPU, comparing to a large bandwidth, the transfer can be capped by your CPU being unable to encrypt (or decrypt in case of the download) the data as fast as your network is capable of transferring them.



      Common SFTP clients cannot distribute the encryption/decryption among CPU cores, so it's actually a capacity of a single CPU core that limits the transfer speed.



      Use the Windows Task manager to see, if one of the cores is utilized to its maximum during the transfer.







    share|improve this answer




















    • 2





      1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

      – Håkan Lindqvist
      May 24 '15 at 11:08






    • 1





      Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

      – htfree
      May 24 '15 at 19:53











    • I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

      – htfree
      May 25 '15 at 21:42











    • I cannot confirm that without seeing a log file. I just tried to show you all the options.

      – Martin Prikryl
      May 26 '15 at 5:15











    • What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

      – htfree
      May 27 '15 at 2:01















    11














    There are three common factors that affect a transfer speed:



    • Bandwidth - An obvious factor that's apparently not your trouble.



    • Network delay/latency - The SFTP is packet oriented-protocol. When downloading, the SFTP client sends a "read" request to the SFTP server, waits for a response, appends returned data to a local file; and repeats, until the end of the file.



      Even if your connection is fast, if the server is far away (or slow), it takes a time for the data to arrive back. If the client spends this time uselessly waiting, your transfer speed will be low.



      Most SFTP clients (including FileZilla and WinSCP) overcome the problem by both requesting a large chunk of the file in each single "read" request and by sending (queuing) multiple "read" requests without waiting for a response to previous. For example WinSCP can request up to 32 chunks for 32 KB each at once, totaling 1 MB (these are defaults). But if there's a big discrepancy between the bandwidth and the network delay, even that 1 MB can be too small to saturate the bandwidth.



      An underlying TCP protocol can suffer a similar problem. So it's not only how the actual SFTP client is efficient, but also how an underlying TCP layer is efficient.



      See also Bandwidth-delay product on Wikipedia.



      I do not think this is your trouble either, at least if you have used the latest version of WinSCP for the tests. There have been some improvements in the recent releases, which allow WinSCP to utilize high-latency connections as efficiently as FileZilla.




    • CPU - The SFTP, being encrypted, is CPU intensive. If you have a relatively slow CPU, comparing to a large bandwidth, the transfer can be capped by your CPU being unable to encrypt (or decrypt in case of the download) the data as fast as your network is capable of transferring them.



      Common SFTP clients cannot distribute the encryption/decryption among CPU cores, so it's actually a capacity of a single CPU core that limits the transfer speed.



      Use the Windows Task manager to see, if one of the cores is utilized to its maximum during the transfer.







    share|improve this answer




















    • 2





      1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

      – Håkan Lindqvist
      May 24 '15 at 11:08






    • 1





      Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

      – htfree
      May 24 '15 at 19:53











    • I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

      – htfree
      May 25 '15 at 21:42











    • I cannot confirm that without seeing a log file. I just tried to show you all the options.

      – Martin Prikryl
      May 26 '15 at 5:15











    • What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

      – htfree
      May 27 '15 at 2:01













    11












    11








    11







    There are three common factors that affect a transfer speed:



    • Bandwidth - An obvious factor that's apparently not your trouble.



    • Network delay/latency - The SFTP is packet oriented-protocol. When downloading, the SFTP client sends a "read" request to the SFTP server, waits for a response, appends returned data to a local file; and repeats, until the end of the file.



      Even if your connection is fast, if the server is far away (or slow), it takes a time for the data to arrive back. If the client spends this time uselessly waiting, your transfer speed will be low.



      Most SFTP clients (including FileZilla and WinSCP) overcome the problem by both requesting a large chunk of the file in each single "read" request and by sending (queuing) multiple "read" requests without waiting for a response to previous. For example WinSCP can request up to 32 chunks for 32 KB each at once, totaling 1 MB (these are defaults). But if there's a big discrepancy between the bandwidth and the network delay, even that 1 MB can be too small to saturate the bandwidth.



      An underlying TCP protocol can suffer a similar problem. So it's not only how the actual SFTP client is efficient, but also how an underlying TCP layer is efficient.



      See also Bandwidth-delay product on Wikipedia.



      I do not think this is your trouble either, at least if you have used the latest version of WinSCP for the tests. There have been some improvements in the recent releases, which allow WinSCP to utilize high-latency connections as efficiently as FileZilla.




    • CPU - The SFTP, being encrypted, is CPU intensive. If you have a relatively slow CPU, comparing to a large bandwidth, the transfer can be capped by your CPU being unable to encrypt (or decrypt in case of the download) the data as fast as your network is capable of transferring them.



      Common SFTP clients cannot distribute the encryption/decryption among CPU cores, so it's actually a capacity of a single CPU core that limits the transfer speed.



      Use the Windows Task manager to see, if one of the cores is utilized to its maximum during the transfer.







    share|improve this answer















    There are three common factors that affect a transfer speed:



    • Bandwidth - An obvious factor that's apparently not your trouble.



    • Network delay/latency - The SFTP is packet oriented-protocol. When downloading, the SFTP client sends a "read" request to the SFTP server, waits for a response, appends returned data to a local file; and repeats, until the end of the file.



      Even if your connection is fast, if the server is far away (or slow), it takes a time for the data to arrive back. If the client spends this time uselessly waiting, your transfer speed will be low.



      Most SFTP clients (including FileZilla and WinSCP) overcome the problem by both requesting a large chunk of the file in each single "read" request and by sending (queuing) multiple "read" requests without waiting for a response to previous. For example WinSCP can request up to 32 chunks for 32 KB each at once, totaling 1 MB (these are defaults). But if there's a big discrepancy between the bandwidth and the network delay, even that 1 MB can be too small to saturate the bandwidth.



      An underlying TCP protocol can suffer a similar problem. So it's not only how the actual SFTP client is efficient, but also how an underlying TCP layer is efficient.



      See also Bandwidth-delay product on Wikipedia.



      I do not think this is your trouble either, at least if you have used the latest version of WinSCP for the tests. There have been some improvements in the recent releases, which allow WinSCP to utilize high-latency connections as efficiently as FileZilla.




    • CPU - The SFTP, being encrypted, is CPU intensive. If you have a relatively slow CPU, comparing to a large bandwidth, the transfer can be capped by your CPU being unable to encrypt (or decrypt in case of the download) the data as fast as your network is capable of transferring them.



      Common SFTP clients cannot distribute the encryption/decryption among CPU cores, so it's actually a capacity of a single CPU core that limits the transfer speed.



      Use the Windows Task manager to see, if one of the cores is utilized to its maximum during the transfer.








    share|improve this answer














    share|improve this answer



    share|improve this answer








    edited Jan 18 '18 at 10:37

























    answered May 24 '15 at 7:07









    Martin PrikrylMartin Prikryl

    5,3102559




    5,3102559







    • 2





      1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

      – Håkan Lindqvist
      May 24 '15 at 11:08






    • 1





      Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

      – htfree
      May 24 '15 at 19:53











    • I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

      – htfree
      May 25 '15 at 21:42











    • I cannot confirm that without seeing a log file. I just tried to show you all the options.

      – Martin Prikryl
      May 26 '15 at 5:15











    • What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

      – htfree
      May 27 '15 at 2:01












    • 2





      1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

      – Håkan Lindqvist
      May 24 '15 at 11:08






    • 1





      Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

      – htfree
      May 24 '15 at 19:53











    • I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

      – htfree
      May 25 '15 at 21:42











    • I cannot confirm that without seeing a log file. I just tried to show you all the options.

      – Martin Prikryl
      May 26 '15 at 5:15











    • What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

      – htfree
      May 27 '15 at 2:01







    2




    2





    1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

    – Håkan Lindqvist
    May 24 '15 at 11:08





    1.3 MB/s is not a whole lot, you'd have to either already be under load for other reasons or be on the extreme low end for CPU to be a limiting factor already with that little throughput. Without knowing all the details, I think BDP seems like the most likely candidate.

    – Håkan Lindqvist
    May 24 '15 at 11:08




    1




    1





    Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

    – htfree
    May 24 '15 at 19:53





    Thanks for all the info. Also, my total cpu load is going from just 2% to 8% during a transfer, utilizing seems 4 cores. (I have an i7 and with hyperthreading so gives 8) I tried with latest version winscp 5.73 and with sftp mode I got 555kb/s and about 805kb/s max with scp mode. Whereas if I start a secondary concurrent transfer in Filezilla I get constant 1.3MiB/s for it also. Ping to the server I'm downloading from is a pretty constant between 180ms to 190ms.

    – htfree
    May 24 '15 at 19:53













    I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

    – htfree
    May 25 '15 at 21:42





    I would mark as accepted answer after confirmation/affirmation if the bandwidth delay product is the issue due to my 180 to 190ms ping delay.

    – htfree
    May 25 '15 at 21:42













    I cannot confirm that without seeing a log file. I just tried to show you all the options.

    – Martin Prikryl
    May 26 '15 at 5:15





    I cannot confirm that without seeing a log file. I just tried to show you all the options.

    – Martin Prikryl
    May 26 '15 at 5:15













    What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

    – htfree
    May 27 '15 at 2:01





    What logfile, I was hoping the additional info such as constant 186ms ping and winscp latest version numbers might shed some more light since I've completely ruled out any CPU bottleneck. So I'm wondering if 180-190ms ping is my limiting factor for the 1.3MiB/s filezilla sftp download or not. I've clicked up your answer as useful but as it stands I can't yet close this question as answered. thanks

    – htfree
    May 27 '15 at 2:01













    2














    I had this issue as well.



    I Used task manager to set the priority to high.



    Now I get up to 5 MiB/s






    share|improve this answer























    • I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

      – htfree
      Jul 23 '16 at 3:32















    2














    I had this issue as well.



    I Used task manager to set the priority to high.



    Now I get up to 5 MiB/s






    share|improve this answer























    • I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

      – htfree
      Jul 23 '16 at 3:32













    2












    2








    2







    I had this issue as well.



    I Used task manager to set the priority to high.



    Now I get up to 5 MiB/s






    share|improve this answer













    I had this issue as well.



    I Used task manager to set the priority to high.



    Now I get up to 5 MiB/s







    share|improve this answer












    share|improve this answer



    share|improve this answer










    answered Mar 22 '16 at 12:10









    Dan Warden DonnellyDan Warden Donnelly

    211




    211












    • I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

      – htfree
      Jul 23 '16 at 3:32

















    • I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

      – htfree
      Jul 23 '16 at 3:32
















    I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

    – htfree
    Jul 23 '16 at 3:32





    I think it helped me a bit also, I also switched to FTPS mode instead of ssh sftp which helps also seems.

    – htfree
    Jul 23 '16 at 3:32











    0














    I recently tried on same exact network with windows 10 and perhaps newer version of filezilla and I got up to 7MB/second transfers from the same server! I then tested with RSYNC inside a virtual machine and got 7MB/second also. I am "pretty sure" now that the problem lies with the COMODO firewall I have installed on this Windows 7 system.



    Apparently even if you "disable" it, all it does is not enforce rules but it slows down network stack. I have installed/replicated this windows 7 system inside a virtual machine also and I will try to completely "remove" Comodo cis premium (antivirus+firewall) and confirm here. I should also mention on this machine I also noticed erratic intermittent latency pings to some systems on my network where all other systems between that were stable <1ms. So bandwidth delay product info is very good but in my case I was able to get filezilla and rsync both at 7MB/s (which basically saturates my available bandwidth) on another install, same network local and remote.






    share|improve this answer



























      0














      I recently tried on same exact network with windows 10 and perhaps newer version of filezilla and I got up to 7MB/second transfers from the same server! I then tested with RSYNC inside a virtual machine and got 7MB/second also. I am "pretty sure" now that the problem lies with the COMODO firewall I have installed on this Windows 7 system.



      Apparently even if you "disable" it, all it does is not enforce rules but it slows down network stack. I have installed/replicated this windows 7 system inside a virtual machine also and I will try to completely "remove" Comodo cis premium (antivirus+firewall) and confirm here. I should also mention on this machine I also noticed erratic intermittent latency pings to some systems on my network where all other systems between that were stable <1ms. So bandwidth delay product info is very good but in my case I was able to get filezilla and rsync both at 7MB/s (which basically saturates my available bandwidth) on another install, same network local and remote.






      share|improve this answer

























        0












        0








        0







        I recently tried on same exact network with windows 10 and perhaps newer version of filezilla and I got up to 7MB/second transfers from the same server! I then tested with RSYNC inside a virtual machine and got 7MB/second also. I am "pretty sure" now that the problem lies with the COMODO firewall I have installed on this Windows 7 system.



        Apparently even if you "disable" it, all it does is not enforce rules but it slows down network stack. I have installed/replicated this windows 7 system inside a virtual machine also and I will try to completely "remove" Comodo cis premium (antivirus+firewall) and confirm here. I should also mention on this machine I also noticed erratic intermittent latency pings to some systems on my network where all other systems between that were stable <1ms. So bandwidth delay product info is very good but in my case I was able to get filezilla and rsync both at 7MB/s (which basically saturates my available bandwidth) on another install, same network local and remote.






        share|improve this answer













        I recently tried on same exact network with windows 10 and perhaps newer version of filezilla and I got up to 7MB/second transfers from the same server! I then tested with RSYNC inside a virtual machine and got 7MB/second also. I am "pretty sure" now that the problem lies with the COMODO firewall I have installed on this Windows 7 system.



        Apparently even if you "disable" it, all it does is not enforce rules but it slows down network stack. I have installed/replicated this windows 7 system inside a virtual machine also and I will try to completely "remove" Comodo cis premium (antivirus+firewall) and confirm here. I should also mention on this machine I also noticed erratic intermittent latency pings to some systems on my network where all other systems between that were stable <1ms. So bandwidth delay product info is very good but in my case I was able to get filezilla and rsync both at 7MB/s (which basically saturates my available bandwidth) on another install, same network local and remote.







        share|improve this answer












        share|improve this answer



        share|improve this answer










        answered Apr 25 at 2:50









        htfreehtfree

        1363616




        1363616



























            draft saved

            draft discarded
















































            Thanks for contributing an answer to Server Fault!


            • Please be sure to answer the question. Provide details and share your research!

            But avoid


            • Asking for help, clarification, or responding to other answers.

            • Making statements based on opinion; back them up with references or personal experience.

            To learn more, see our tips on writing great answers.




            draft saved


            draft discarded














            StackExchange.ready(
            function ()
            StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fserverfault.com%2fquestions%2f694062%2fwhy-is-filezilla-sftp-file-transfer-max-capped-at-1-3mib-sec-instead-of-saturati%23new-answer', 'question_page');

            );

            Post as a guest















            Required, but never shown





















































            Required, but never shown














            Required, but never shown












            Required, but never shown







            Required, but never shown

































            Required, but never shown














            Required, but never shown












            Required, but never shown







            Required, but never shown







            Popular posts from this blog

            Club Baloncesto Breogán Índice Historia | Pavillón | Nome | O Breogán na cultura popular | Xogadores | Adestradores | Presidentes | Palmarés | Historial | Líderes | Notas | Véxase tamén | Menú de navegacióncbbreogan.galCadroGuía oficial da ACB 2009-10, páxina 201Guía oficial ACB 1992, páxina 183. Editorial DB.É de 6.500 espectadores sentados axeitándose á última normativa"Estudiantes Junior, entre as mellores canteiras"o orixinalHemeroteca El Mundo Deportivo, 16 setembro de 1970, páxina 12Historia do BreogánAlfredo Pérez, o último canoneiroHistoria C.B. BreogánHemeroteca de El Mundo DeportivoJimmy Wright, norteamericano do Breogán deixará Lugo por ameazas de morteResultados de Breogán en 1986-87Resultados de Breogán en 1990-91Ficha de Velimir Perasović en acb.comResultados de Breogán en 1994-95Breogán arrasa al Barça. "El Mundo Deportivo", 27 de setembro de 1999, páxina 58CB Breogán - FC BarcelonaA FEB invita a participar nunha nova Liga EuropeaCharlie Bell na prensa estatalMáximos anotadores 2005Tempada 2005-06 : Tódolos Xogadores da Xornada""Non quero pensar nunha man negra, mais pregúntome que está a pasar""o orixinalRaúl López, orgulloso dos xogadores, presume da boa saúde económica do BreogánJulio González confirma que cesa como presidente del BreogánHomenaxe a Lisardo GómezA tempada do rexurdimento celesteEntrevista a Lisardo GómezEl COB dinamita el Pazo para forzar el quinto (69-73)Cafés Candelas, patrocinador del CB Breogán"Suso Lázare, novo presidente do Breogán"o orixinalCafés Candelas Breogán firma el mayor triunfo de la historiaEl Breogán realizará 17 homenajes por su cincuenta aniversario"O Breogán honra ao seu fundador e primeiro presidente"o orixinalMiguel Giao recibiu a homenaxe do PazoHomenaxe aos primeiros gladiadores celestesO home que nos amosa como ver o Breo co corazónTita Franco será homenaxeada polos #50anosdeBreoJulio Vila recibirá unha homenaxe in memoriam polos #50anosdeBreo"O Breogán homenaxeará aos seus aboados máis veteráns"Pechada ovación a «Capi» Sanmartín e Ricardo «Corazón de González»Homenaxe por décadas de informaciónPaco García volve ao Pazo con motivo do 50 aniversario"Resultados y clasificaciones""O Cafés Candelas Breogán, campión da Copa Princesa""O Cafés Candelas Breogán, equipo ACB"C.B. Breogán"Proxecto social"o orixinal"Centros asociados"o orixinalFicha en imdb.comMario Camus trata la recuperación del amor en 'La vieja música', su última película"Páxina web oficial""Club Baloncesto Breogán""C. B. Breogán S.A.D."eehttp://www.fegaba.com

            Vilaño, A Laracha Índice Patrimonio | Lugares e parroquias | Véxase tamén | Menú de navegación43°14′52″N 8°36′03″O / 43.24775, -8.60070

            Cegueira Índice Epidemioloxía | Deficiencia visual | Tipos de cegueira | Principais causas de cegueira | Tratamento | Técnicas de adaptación e axudas | Vida dos cegos | Primeiros auxilios | Crenzas respecto das persoas cegas | Crenzas das persoas cegas | O neno deficiente visual | Aspectos psicolóxicos da cegueira | Notas | Véxase tamén | Menú de navegación54.054.154.436928256blindnessDicionario da Real Academia GalegaPortal das Palabras"International Standards: Visual Standards — Aspects and Ranges of Vision Loss with Emphasis on Population Surveys.""Visual impairment and blindness""Presentan un plan para previr a cegueira"o orixinalACCDV Associació Catalana de Cecs i Disminuïts Visuals - PMFTrachoma"Effect of gene therapy on visual function in Leber's congenital amaurosis"1844137110.1056/NEJMoa0802268Cans guía - os mellores amigos dos cegosArquivadoEscola de cans guía para cegos en Mortágua, PortugalArquivado"Tecnología para ciegos y deficientes visuales. Recopilación de recursos gratuitos en la Red""Colorino""‘COL.diesis’, escuchar los sonidos del color""COL.diesis: Transforming Colour into Melody and Implementing the Result in a Colour Sensor Device"o orixinal"Sistema de desarrollo de sinestesia color-sonido para invidentes utilizando un protocolo de audio""Enseñanza táctil - geometría y color. Juegos didácticos para niños ciegos y videntes""Sistema Constanz"L'ocupació laboral dels cecs a l'Estat espanyol està pràcticament equiparada a la de les persones amb visió, entrevista amb Pedro ZuritaONCE (Organización Nacional de Cegos de España)Prevención da cegueiraDescrición de deficiencias visuais (Disc@pnet)Braillín, un boneco atractivo para calquera neno, con ou sen discapacidade, que permite familiarizarse co sistema de escritura e lectura brailleAxudas Técnicas36838ID00897494007150-90057129528256DOID:1432HP:0000618D001766C10.597.751.941.162C97109C0155020