1 |
On Fri, 2012-06-22 at 05:10 -0400, Philip Webb wrote: |
2 |
> 120622 Mick wrote: |
3 |
> > On Friday 22 Jun 2012 05:41:14 Philip Webb wrote: |
4 |
> >> In recent weeks, I've seen many instances of downloads of files |
5 |
> >> coming in small segments, which seem to be the same for each server, |
6 |
> >> but vary between different servers. Some Gentoo mirrors suffer this way |
7 |
> >> -- not all -- & some general news sites, eg when delivering videos, |
8 |
> >> which run for a few secs, then freeze, then resume etc. |
9 |
> >> Once I observed a server doing this for several hours (a large file), |
10 |
> >> then suddenly loosening up & delivering the rest of the file in one go. |
11 |
> > Yes, when I use the wireless network at work. |
12 |
> > It could be the ISP or it could be other users taking up bandwidth. |
13 |
> > Really early in the morning performance is generally higher |
14 |
> > than during work hours. this could be due to contention |
15 |
> > on the local network or the ISP's pipe. |
16 |
> > Do you get such problems off peak hours? |
17 |
> |
18 |
> I think you're describing a different phenomenon, traffic jams (smile). |
19 |
> That sometimes happens here & may persist for 1 - 2 weeks , |
20 |
> but it feels just like driving on the highway when 1 lane is closed. |
21 |
> That probably is some piece of the Internet or ISP under repair or test. |
22 |
> |
23 |
> What I'm refering to is getting downloaded files in slices, |
24 |
> eg using Wget, a piece of the file downloads for 22 sec , then stops; |
25 |
> Wget tries again & another 22 sec piece comes down the pipe, then stops. |
26 |
> This can go on for hours with a big file |
27 |
> & doesn't seem related to the local time of day. |
28 |
> My guess is that the server has been programmed to stop after 22 sec |
29 |
> in an effort to share access among many clients, |
30 |
> but it is irritating & also suggests the server needs faster hardware. |
31 |
> |
32 |
> Have others noticed this -- it seems to be a recent innovation -- |
33 |
> & is it a known ploy of server managers ? |
34 |
> |
35 |
|
36 |
Are you using traffic shaping like a "police filter", or is there |
37 |
shaping somewhere in the path? - most of the linux shaping methods work |
38 |
on a burst principle that produces a given throughput by gating the |
39 |
traffic for an average throughput. I have seen the effect you mention |
40 |
with block transfer protocols (ftp) and wget when trying to use its |
41 |
built in bandwidth regulation which works similarly to when traffic |
42 |
shaping is in effect. There seems to be an interaction between the two |
43 |
instances of regulating the traffic to create an effect like you have |
44 |
seen. This also affects other traffic trying to use the link at the |
45 |
same time as the available bandwidth gets very "choppy". |
46 |
|
47 |
I have less experience with other than linux based shaping methods such |
48 |
as Cisco QoS methods but have not seen the same effect there. |
49 |
|
50 |
BillK |