[tahoe-dev] Tahoe-LAFS + sshfs latency issues
Zooko O'Whielacronx
zooko at zooko.com
Fri Jul 22 23:36:32 PDT 2011
On Fri, Jul 22, 2011 at 7:10 PM, Linus Lüssing <linus.luessing at web.de> wrote:
> Hi everyone,
>
> I've recently started looking at Tahoe-LAFS and I'm currently trying to get a feeling of how it works.
Welcome, Linus!
> The use-case I'm currently aiming at is to have an ftpd and httpd (proftpd + lighttpd) running on a distributed file system, where anyone can just add and contribute to the storage capacity.
Cool!
> Furthermore I'm going to use it within an open, public, decentral wireless+vpn mesh network (e.g. using the mesh routing protocol B.A.T.M.A.N.-Advanced and the VPN software tinc).
Cool!
I look forward to hearing about how well this works.
> time scp -P 8022 root at 127.0.0.1:/music/8bitpeoples/8BP102.gif /tmp/:
> 0m1.050s, 0m0.838s, 0m0.893s
>
> time cp /home/ftp/music/8bitpeoples/8BP102.gif /tmp/:
> 0m25.539s, 0m29.389s, 0m27.414s
>
>
> Without the 'tahoe put' running everything works as expected:
>
> time scp -P 8022 root at 127.0.0.1:/music/8bitpeoples/8BP102.gif /tmp/:
> 0m0.937s, 0m1.001s, 0m0.962s
>
> time cp /home/ftp/music/8bitpeoples/8BP102.gif /tmp/
> 0m0.651s, 0m0.587s, 0m0.624s
Okay let me get this straight. The scp is using Tahoe-LAFS SFTPD which
is listening on port 8022, right? And the cp is using sshfs which is
using that same Tahoe-LAFS SFTPD. So any difference in behavior
between those two is either due to sshfs itself rather than to
Tahoe-LAFS, or to some difference in the way sshfs uses the SFTPD vs.
the way that scp uses the SFTPD, right?
And what you observe is that whenever you are doing a lot of uploads
to the Tahoe-LAFS gateway then the cp takes 25 to 30 times longer?
Wow!
This is an interesting bug. Are you sure it is reproducible?
Okay, two next steps you could do:
1. Look for logs. We've written some notes on how to find them in "How
To Report A Bug" [1], and thorough details about logging in
"docs/logging.rst" [2].
2. Look on the front page of the tahoe-lafs gateway for the Recent
Uploads and Downloads page. From there you can get performance
measurements of all (*) uploads and downloads.
Thank you very much for investigating and reporting this issue.
Regards,
Zooko
P.S. The fact that I have to tell you about the existence of "Recent
Uploads and Downloads" is because of bug #1375.
[1] http://tahoe-lafs.org/trac/tahoe-lafs/wiki/HowToReportABug
[2] http://tahoe-lafs.org/trac/tahoe-lafs/browser/trunk/docs/logging.rst
(*) Except for the ones that are omitted due to a bug (#1079).
http://tahoe-lafs.org/trac/tahoe-lafs/ticket/1079# upload of file into
dir doesn't appear on Recent Uploads and Downloads
http://tahoe-lafs.org/trac/tahoe-lafs/ticket/1375# the performance
stats for each upload or download are really well hidden
More information about the tahoe-dev
mailing list