We are experimenting with a mirror that seems to work good.

I mailed you some info privately.

I am not sure if we can make this public already (Marcus ?)

> On 17 Nov 2017, at 13:05, Ben Coman <[email protected]> wrote:
> 
> Performance has been fairly reasonable lately, until just now...
> from PharoLauncher I downloaded Pharo 7 latest-32 and that zipped along okay,
> then immediately downloaded Pharo 7 latest-64 and it crawled.  About 30 
> seconds or more per 1%. That would have been ~16:50TZ+8)
> I went to dinner.  After dinner ~19:30TZ+8 I found it timed out and kicked 
> off another that was a tiny bit better, but still crawling.
> Interestingly, from a terminal I then did... $ wget 
> http://files.pharo.org/image/70/latest-64.zip
> and for the first 15% it crawled at 20kB/s. Then up to 50% it ramped steadily 
> up to 400kB/s, then the last 50% it smoothly ramped up 1.5MB/s and by the end 
> was zooming.  Meanwhile, PharoLauncher continued plodding along and finished 
> a few minutes later with a "CRCError: Wrong CRC-32 (expected 55B3346A got 
> 2DD68EFA) (proceed to ignore)" , which I've never seen before.
> 
> About five minutes later I tried wget with the same behaviour, and then 
> PharoLauncher again and it was much better taking about 20 seconds to 
> successfully download Pharo 7 latest-64.zip.  Paying close attention the 
> download speed of PharoLauncher seemed to increase *slightly* as it went 
> along, but nothing like wget's progressive improvement. 
> 
> I'm not quite sure what to make of that.  The overlapped downloads by 
> PharoLauncher and wget performing differently seems significant.  Does the 
> server have some throttling mechanism that maybe adapts its policies due to 
> overall load, which Pharo is interacting with poorly?  
> 
> 
> Another observation, using Chromium I found pages [1] 
> http://files.pharo.org/image/60/ and [2] http://files.pharo.org/image/70/ 
> very slow, with the latter being worse.  In one instance refreshing [2] I 
> counted 40 seconds with no response (blank page).  Then tried refreshing [2] 
> again and at the the 6 second mark refreshed [1], then at a count of 20 both 
> pages simultaneously appeared.  
> 
> No other internet services are slow.  My current connection speedtest rates 
> my link at 34Mbps (~4MB/s).
> 
> Unfortunately I didn't get a traceroute while things were really slow.  Here 
> is one after it got back to normal (so the poor performance lasted about 50 
> minutes).
> 
> $ traceroute files.pharo.org
> traceroute to files.pharo.org (128.93.162.72), 30 hops max, 60 byte packets
>  1  192.168.43.1 (192.168.43.1)  3.122 ms  3.180 ms  4.437 ms
>  2  * * *
>  3  * * *
>  4  Bundle-Ether14.wel-edge901.perth.telstra.net (139.130.16.49)  70.991 ms  
> 71.000 ms  70.984 ms
>  5  bundle-ether6.wel-core3.perth.telstra.net (203.50.6.214)  70.964 ms  
> 70.944 ms  72.089 ms
>  6  bundle-ether7.fli-core1.adelaide.telstra.net (203.50.11.18)  85.918 ms  
> 73.619 ms  82.271 ms
>  7  bundle-ether9.win-core10.melbourne.telstra.net (203.50.11.91)  98.565 ms  
> 98.568 ms  98.560 ms
>  8  bundle-ether12.ken-core10.sydney.telstra.net (203.50.11.122)  109.279 ms  
> 109.298 ms  109.288 ms
>  9  bundle-ether1.pad-gw11.sydney.telstra.net (203.50.6.61)  109.232 ms  
> 98.702 ms  101.279 ms
> 10  bundle-ether1.sydp-core04.sydney.reach.com (203.50.13.90)  92.535 ms  
> 104.351 ms  113.158 ms
> 11  i-52.tlot-core02.bx.telstraglobal.net (202.84.137.101)  294.263 ms  
> 283.547 ms  281.576 ms
> 12  i-0-7-0-13.tlot-core01.bi.telstraglobal.net (202.40.149.165)  279.967 ms 
> i-0-2-0-11.tlot-core01.bi.telstraglobal.net (202.40.149.185)  279.948 ms 
> i-0-7-0-11.tlot-core01.bi.telstraglobal.net (202.84.251.233)  277.390 ms
> 13  gtt-peer.tlot02.pr.telstraglobal.net (134.159.63.182)  280.217 ms  
> 288.853 ms  275.764 ms
> 14  xe-1-2-0.cr0-par7.ip4.gtt.net (89.149.185.53)  461.558 ms  445.153 ms  
> 445.097 ms
> 15  renater-gw-ix1.gtt.net (77.67.123.206)  445.083 ms  443.652 ms  441.766 ms
> 16  * * *
> 17  inria-rocquencourt-te1-4-inria-rtr-021.noc.renater.fr (193.51.184.177)  
> 421.283 ms  480.646 ms  424.906 ms
> 
>  
> 
> 
> 
> cheers -ben
> 
> On 29 September 2017 at 03:11, Stephane Ducasse <[email protected]> 
> wrote:
> We opened a new ticket.
> 
> 
> On Tue, Sep 26, 2017 at 8:45 PM, Gabriel Cotelli <[email protected]> wrote:
> > I'm willing to do some other specific tests if needed. Just let me know.
> >
> > On Tue, Sep 26, 2017 at 3:36 PM, Stephane Ducasse <[email protected]>
> > wrote:
> >>
> >> Thanks Gabriel I forwarded this to christophe.
> >>
> >> On Tue, Sep 26, 2017 at 12:01 AM, Gabriel Cotelli <[email protected]>
> >> wrote:
> >> > Still really slow from Argentina.
> >> >
> >> > In case it helps
> >> >
> >> > traceroute  files.pharo.org
> >> > traceroute to files.pharo.org (128.93.162.72), 30 hops max, 60 byte
> >> > packets
> >> >  1  192.168.1.1 (192.168.1.1)  0.574 ms  0.968 ms  1.538 ms
> >> >  2  192.168.0.1 (192.168.0.1)  2.048 ms  2.499 ms  2.847 ms
> >> >  3  10.33.32.1 (10.33.32.1)  12.264 ms  14.160 ms  14.336 ms
> >> >  4  10.242.2.161 (10.242.2.161)  14.522 ms  14.643 ms  16.497 ms
> >> >  5  telecentro.baires3.bai.seabone.net (195.22.220.33)  18.936 ms
> >> > 19.575 ms
> >> > 18.843 ms
> >> >  6  ae0.baires3.bai.seabone.net (195.22.220.32)  19.554 ms  10.675 ms
> >> > 13.028 ms
> >> >  7  et10-1-0.miami15.mia.seabone.net (89.221.41.175)  246.284 ms
> >> > et9-3-0.miami15.mia.seabone.net (195.22.199.179)  246.214 ms
> >> > et4-3-0.miami15.mia.seabone.net (195.22.199.177)  246.137 ms
> >> >  8  gtt.miami15.mia.seabone.net (89.221.41.197)  245.909 ms  245.924 ms
> >> > 245.870 ms
> >> >  9  xe-3-1-2.cr0-par7.ip4.gtt.net (141.136.109.18)  251.632 ms  274.447
> >> > ms
> >> > 250.097 ms
> >> > 10  renater-gw-ix1.gtt.net (77.67.123.206)  281.025 ms  259.509 ms
> >> > 261.717
> >> > ms
> >> > 11  * * *
> >> > 12  inria-rocquencourt-te1-4-inria-rtr-021.noc.renater.fr
> >> > (193.51.184.177)
> >> > 271.263 ms  307.011 ms  306.876 ms
> >> > 13  * * *
> >> > 14  * * *
> >> > 15  * * *
> >> > 16  * * *
> >> > 17  * * *
> >> > 18  * * *
> >> > 19  * * *
> >> > 20  * * *
> >> > 21  * * *
> >> > 22  * * *
> >> > 23  * * *
> >> > 24  * * *
> >> > 25  * * *
> >> > 26  * * *
> >> > 27  * * *
> >> > 28  * * *
> >> > 29  * * *
> >> > 30  * * *
> >> >
> >> > httping  -g http://files.pharo.org/get-files/61/pharo64.zip -c 20
> >> > PING files.pharo.org:80 (/get-files/61/pharo64.zip):
> >> > connected to 128.93.162.72:80 (256 bytes), seq=0 time=625,90 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=1 time=842,02 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=2 time=638,83 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=3 time=637,63 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=4 time=637,85 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=5 time=580,27 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=6 time=1207,98 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=7 time=638,13 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=8 time=637,74 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=9 time=582,59 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=10 time=658,11 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=11 time=673,84 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=12 time=584,44 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=13 time=691,84 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=14 time=637,20 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=15 time=861,84 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=16 time=619,45 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=17 time=638,03 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=18 time=638,11 ms
> >> > connected to 128.93.162.72:80 (256 bytes), seq=19 time=741,01 ms
> >> > --- http://files.pharo.org/get-files/61/pharo64.zip ping statistics ---
> >> > 20 connects, 20 ok, 0,00% failed, time 33781ms
> >> > round-trip min/avg/max = 580,3/688,6/1208,0 ms
> >> >
> >> >
> >> > On Thu, Sep 21, 2017 at 12:13 PM, Ben Coman <[email protected]> wrote:
> >> >>
> >> >> Still seems a bit slow...
> >> >> http://files.pharo.org/get-files/70/pharo64.zip
> >> >> 16MB ==> 10 minutes (and actually it stalled after 3.4MB downloaded)
> >> >> Local speed test: 40Mb/s down, 20Mb/s up
> >> >>
> >> >> but then I'm a long way across the internet, the cause could be
> >> >> anywhere...
> >> >> $ traceroute files.pharo.org
> >> >> traceroute to files.pharo.org (128.93.162.72), 30 hops max, 60 byte
> >> >> packets
> >> >>  1  192.168.43.1 (192.168.43.1)  2.040 ms  2.403 ms  4.222 ms
> >> >>  2  * * *
> >> >>  3  * * *
> >> >>  4  Bundle-Ether14.wel-edge901.perth.telstra.net (139.130.16.49)
> >> >> 67.741
> >> >> ms  67.739 ms  67.730 ms
> >> >>  5  bundle-ether6.wel-core3.perth.telstra.net (203.50.6.214)  67.696 ms
> >> >> 67.697 ms  67.697 ms
> >> >>  6  bundle-ether7.fli-core1.adelaide.telstra.net (203.50.11.18)  95.723
> >> >> ms
> >> >> 91.753 ms  100.574 ms
> >> >>  7  bundle-ether9.win-core10.melbourne.telstra.net (203.50.11.91)
> >> >> 106.550
> >> >> ms  108.849 ms  108.851 ms
> >> >>  8  bundle-ether12.ken-core10.sydney.telstra.net (203.50.11.122)
> >> >> 117.770
> >> >> ms  117.767 ms  117.751 ms
> >> >>  9  bundle-ether1.pad-gw11.sydney.telstra.net (203.50.6.61)  117.720 ms
> >> >> 103.603 ms  102.171 ms
> >> >> 10  bundle-ether1.sydp-core04.sydney.reach.com (203.50.13.90)  93.276
> >> >> ms
> >> >> 95.822 ms  102.577 ms
> >> >> 11  i-52.tlot-core02.bx.telstraglobal.net (202.84.137.101)  330.128 ms
> >> >> 321.231 ms  322.522 ms
> >> >> 12  i-0-4-0-10.tlot-core01.bi.telstraglobal.net (202.84.253.21)
> >> >> 321.187
> >> >> ms i-0-2-0-10.tlot-core01.bi.telstraglobal.net (202.40.149.181)
> >> >> 321.181 ms
> >> >> i-0-2-0-9.tlot-core01.bi.telstraglobal.net (202.40.149.177)  308.786 ms
> >> >> 13  gtt-peer.tlot02.pr.telstraglobal.net (134.159.63.182)  273.784 ms
> >> >> 283.530 ms  281.079 ms
> >> >> 14  xe-1-2-0.cr0-par7.ip4.gtt.net (89.149.185.53)  466.648 ms  479.910
> >> >> ms
> >> >> 481.254 ms
> >> >> 15  renater-gw-ix1.gtt.net (77.67.123.206)  479.900 ms  479.882 ms
> >> >> 478.641 ms
> >> >> 16  * * *
> >> >> 17  inria-rocquencourt-te1-4-inria-rtr-021.noc.renater.fr
> >> >> (193.51.184.177)
> >> >> 436.776 ms  479.819 ms  478.907 ms
> >> >> 18  * * *
> >> >> 19  * * *
> >> >>
> >> >> $ httping -h files.pharo.org
> >> >> PING files.pharo.org:80 (http://files.pharo.org:80/):
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=0 time=970.27 ms
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=1 time=928.26 ms
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=2 time=860.06 ms
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=3 time=976.04 ms
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=4 time=1238.72 ms
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=5 time=1239.65 ms
> >> >> connected to 128.93.162.72:80 (169 bytes), seq=6 time=918.06 ms
> >> >>
> >> >>
> >> >> cheers -ben
> >> >>
> >> >> On Thu, Sep 21, 2017 at 5:43 PM, Marcus Denker <[email protected]>
> >> >> wrote:
> >> >>>
> >> >>>
> >> >>> > On 18 Sep 2017, at 08:37, Marcus Denker <[email protected]>
> >> >>> > wrote:
> >> >>> >
> >> >>> >
> >> >>> >> On 16 Sep 2017, at 09:40, Stephane Ducasse
> >> >>> >> <[email protected]>
> >> >>> >> wrote:
> >> >>> >>
> >> >>> >> Hi marcus
> >> >>> >>
> >> >>> >> I do not see why they could not find it.
> >> >>> >> Can you tell me how you ask because we should reask?
> >> >>> >>
> >> >>> >>
> >> >>> >
> >> >>> > I spend a ticket and they wanted to close it last week. I have told
> >> >>> > them to not close it
> >> >>> >
> >> >>> The problems seems to be fixed now.
> >> >>>
> >> >>> If someone sees any slowness -> send me a mail.
> >> >>>
> >> >>>         Marcus
> >> >>>
> >> >>>
> >> >>
> >> >
> >>
> >
> 
> 


Reply via email to