415933 - Very slow downloads from ftp.m.o

Status: RESOLVED INVALID

(mozilla.org Graveyard :: Server Operations, task)

Description

[Nick Thomas [:nthomas] (UTC+12)]

I get about 15KB/s on file downloads from ftp.m.o, stage.m.o, and localgho.st (a VM on landfill), while connecting to mozilla.ussg.indiana.edu get > 400KB/s. Traceroute to ftp.m.o gives a nominal 170ms latency to the border.

Have also heard reports from nightly testers and the Canadian contingent that nightly updates are slow to d/l (they come off ftp.m.o).

Perhaps some fallout from the maintenance window last night ? Or people are going nuts for the 3.0b3 candidate builds ? The latter would mean the ftp daemon on ftp.m.o was very busy (http is redirected).

Comment 1

[Nick Thomas [:nthomas] (UTC+12)]

(In reply to comment #0)
> I get about 15KB/s on file downloads from ftp.m.o, stage.m.o, and localgho.st
> (a VM on landfill), while connecting to mozilla.ussg.indiana.edu get > 400KB/s.
> Traceroute to ftp.m.o gives a nominal 170ms latency to the border.

Forgot to say that about 70KB/s is normal at this time of day.

Comment 2

[Justin Fitzhugh]

Just downloaded b3 to my house in the SF bay area on comcast at over 1mb/s - don't think it's a server issue.  Could be a upstream transit issue for international people...

Comment 3

[matthew zeier [:mrz]]

I agree with justin - works for me off comcast too and appears quick from Amsterdam, about as far away as I can get.  

Perhaps it was transient but in the case, it's even harder to debug now.

Comment 4

[Peter van der Woude [:Peter6]]

this is definitly not WFM from Amsterdam(NL).
So this bug should either be resolved INVALID (not a problem at Mozilla's end) or left NEW.

Comment 5

[matthew zeier [:mrz]]

Hurm, it actually did wfm though. 

Locally within the datacenter I get 9.75M/sec.  Through Comcast I get 600K/second.  From China I get 200K/second (!!).  From Amsterdam I'm seeing about 20K/second (which while sub normal isn't anything I change).  

I'll close it INVALID as it doesn't appear to be a problem on Mozilla's side.

Comment 6

[Nick Thomas [:nthomas] (UTC+12)]

Whatever the routing/provider problem was, it's gone away for me today.