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 > >> >>> > >> >>> > >> >> > >> > > >> > > > >
