[Mike B]

Original: 2156Z: I'm trying to update OR. Also trying to download the MG version. I have (theoretically) 50 mbit connection. Other things (including opening ET) move at the normal rate (nowhere near 50 mbit, but reasonable). But OR updater is very nearly frozen (has reached about 1/4 completion after over 20 minutes), and downloading either OR or OR-MG proceeds at a Firefox-reported 20-25 kbps. Yes, dialup speed. Have the OR download sites switched to Hayes modems, or is something else going on?

Happy Friday? :curiousPC:

Update at 2207Z: MG completed downloading - but it's only 12 MB. OR-Testing from the official web site is still going, about 1/2 way, up to 50kb/sec occasionally. Updater has been running for over 1/2 hour and it's not quite at 1/2 yet. Yes, I know, gripe gripe, but I normally get considerably faster downloads.

[snark]Could it be that the US internet connection with the rest of the world is going through quarantine?[/snark]

Update at 2220Z: Updater has made it past 1/2! OR-Testing has made it to about 75%. Here are some stats I pulled:
Pinging openrails.org [87.237.69.234] with 32 bytes of data:
Request timed out.
Request timed out.
Reply from 87.237.69.234: bytes=32 time=2559ms TTL=109
Reply from 87.237.69.234: bytes=32 time=157ms TTL=109

Ping statistics for 87.237.69.234:
Packets: Sent = 4, Received = 2, Lost = 2 (50% loss),
Approximate round trip times in milli-seconds:
Minimum = 157ms, Maximum = 2559ms, Average = 1358ms

And it looks like the problem is somewhere between AT&T and London (so it's not OR's fault, as if it ever was):
Tracing route to openrails.org [87.237.69.234]
over a maximum of 30 hops:

1 3 ms 3 ms 3 ms dsldevice.attlocal.net [redacted local network IP]
2 24 ms 23 ms 26 ms 75-53-84-1.lightspeed.flsmca.sbcglobal.net [75.53.84.1]
3 24 ms 52 ms 51 ms 71.147.198.21
4 31 ms 30 ms 30 ms 12.122.160.154
5 29 ms 29 ms 26 ms 12.122.2.78
6 28 ms 27 ms 29 ms 12.122.114.5
7 * * * Request timed out.
8 * * * Request timed out.
9 * * * Request timed out.
10 * 1859 ms * HUB-NETWORK.ear2.London15.Level3.net [212.187.173.118]
11 2213 ms 2459 ms 563 ms sv167.lon-thn1.hns.net [31.210.132.49]
12 424 ms * 1273 ms sv7.unicorn1.hns.net [31.210.132.9]
13 158 ms 158 ms 157 ms sv170.unicorn0.hns.net [31.210.132.53]
14 658 ms 164 ms 879 ms sv165.ggs1.hns.net [31.210.132.41]
15 159 ms 160 ms 309 ms 87-237.69-234.hns.net [87.237.69.234]

Trace complete.

Ah well, it's Friday, I need to be patient... :Neeeedsleeep:

[xavivilla]

Downloaded OpenRails-Testing in 30 seconds from Spain

[Mike B]

xavivilla, on 17 July 2020 - 02:11 PM, said:

Downloaded OpenRails-Testing in 30 seconds from Spain

Thanks. So it is just me (or AT&T throttling the download for some reason).

[ErickC]

I had problems with a whole bunch of websites earlier today around the same time, although I'm on Comcast.

[ebnertra000]

Probably AT&T. In my neck of the woods, AT&T works either in cities (spotty) or the middle of nowhere (also spotty). Rural or small town service is possible, but not exactly betting odds.

That said, one of my co-workers said his mobile internet went down (he uses Verizon), and the timesheet software at work went down, as well. Everything was wonky, I guess. Solar flare? Hax? That, or the internet isn't safe from pandemics, either...

[alteo80]

https://www.zerohedg...-suffers-outage

[Genma Saotome]

I reqested our host not to put us on Cloudflare so their problem does not directly affect us. It may affect everything anyway.

FWIW Cloudflair excels at improving response time for relatively static web sites, It does this by replicating web sites to many host locations around the world, working on the premise that the replication will but sites closer to end users. Generally it works. Where it doesn't help is with relatively dynamic web sites such as ours. In our case a query to a Cloudflare site would find whatever they had was not current and so they'd query our original database. In effect this puts an extra layer of processing between end users and the web pages they want to see. I saw nothing good in that prospect so i instructed our host to not put us on Cloudflare and AFGAIK they have not.

[Mike B]

I just saw a report from Techcrunch similar to alteo80's link. So a Cloudflare router in Atlanta took down the internet...

https://techcrunch.c...ternet-with-it/

I tried their "1.1.1.1 DNS and WARP" app on my phone and backed off from it a while back. It started blocking nearly every web site I tried to use and a lot of apps (early warning of the bad router?). Using only the DNS was a little less obtrusive, but still caused problems with some government sites of all things. So it's turned off for now; if I need a VPN on the phone I'll get a real one.

Given the timing and effects, it probably was at least partly responsible for the problems I was having with OR downloads.

[waivethefive]

I downloaded the testing version 1.3.1.191 dated 7-17 from the website just fine, but when I use that version, I am being prompted to "Upgrade" backwards to 1.3.1.189. If I upgrade, it gets into a never ending cycle to upgrade between 2 older vaersions.

If I am on testing version 1.3.1.191 dated 7-17 and switch updater options to unstable, it prompts me to "upgrade" to U2020.7.15-1806 which again is attempting to backdate from where I am.