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 <stepharo.self@gmail.com> wrote:
We opened a new ticket.


On Tue, Sep 26, 2017 at 8:45 PM, Gabriel Cotelli <g.cotelli@gmail.com> 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 <stepharo.self@gmail.com>
> wrote:
>>
>> Thanks Gabriel I forwarded this to christophe.
>>
>> On Tue, Sep 26, 2017 at 12:01 AM, Gabriel Cotelli <g.cotelli@gmail.com>
>> 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 <btc@openinworld.com> 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 <marcus.denker@inria.fr>
>> >> wrote:
>> >>>
>> >>>
>> >>> > On 18 Sep 2017, at 08:37, Marcus Denker <marcus.denker@inria.fr>
>> >>> > wrote:
>> >>> >
>> >>> >
>> >>> >> On 16 Sep 2017, at 09:40, Stephane Ducasse
>> >>> >> <stepharo.self@gmail.com>
>> >>> >> 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
>> >>>
>> >>>
>> >>
>> >
>>
>