Re: Bots, missing nodes and outages

[Corion]

No, the recent outages were due to network issues.

[choroba]

I can't connect from home, but when I run a remote desktop at work, I can connect from there.

traceroute from work:

  1   # Redacted
  2   195.113.69.5  0.366ms  0.307ms  0.289ms 
  3   195.113.69.169  0.360ms  0.299ms  0.343ms 
  4   195.113.67.5  0.462ms  0.445ms  0.458ms 
  5   195.113.69.54  1.265ms  1.079ms  1.081ms 
  6   *  *  * 
  7   80.249.212.170  14.305ms  14.304ms  14.158ms 
  8   *  *  * 
  9   *  *  * 
 10   *  *  * 
 11   *  *  * 
 12   *  *  * 
 13   *  *  * 
 14   *  *  * 
 15   204.16.241.181  128.520ms  116.818ms  118.734ms 
 16   204.16.243.217  122.314ms  113.822ms  111.790ms 
 17   *  *  *  
 18   66.39.54.27  137.129ms  128.118ms  109.758ms

traceroute from home:

 1  192.168.100.1 (192.168.100.1)  1.701 ms  1.659 ms  1.639 ms
 2  10.21.32.1 (10.21.32.1)  9.593 ms  9.574 ms  9.546 ms
 3  switch-bb-pha-stodulky-switch-gpon-pha-stodulky.ptp.poda.network (
+10.11.185.25)  9.462 ms  9.443 ms  9.425 ms
 4  static-3248824331.poda.cz (193.165.32.11)  9.560 ms  9.452 ms  9.5
+23 ms
 5  static-3248824321.poda.cz (193.165.32.1)  9.350 ms  9.331 ms  9.37
+8 ms
 6  datacamp-fra-2.peering.cz (185.0.20.200)  11.132 ms datacamp-fra-1
+.peering.cz (185.0.20.28)  11.202 ms datacamp-fra-2.peering.cz (185.0
+.20.200)  11.174 ms
 7  unn-169-150-192-197.cdn77.com (169.150.192.197)  14.477 ms unn-169
+-150-192-199.cdn77.com (169.150.192.199)  17.906 ms  17.876 ms
 8  * * *
 9  204.16.241.168 (204.16.241.168)  108.787 ms * *
10  204.16.241.183 (204.16.241.183)  108.113 ms *  108.694 ms
11  204.16.243.219 (204.16.243.219)  108.675 ms * *
12  * * *
13  * 204.16.243.219 (204.16.243.219)  109.318 ms *
14  * * *
15  * * *
16  * * *
17  * * *
18  * * *
19  * * *
20  * * *
21  * * *
22  * * *
23  * * *
24  * * *
25  * * *
26  * * *
27  * * *
28  * * *
29  * * *
30  * * *

So, 204.16.243.217 knows the way, but 204.16.243.219 doesn't. whois returns exactly the same details about both the addresses.

[pryrt]

For me, at $work: on Friday, my experiments agreed with choroba's traceroutes above, where 204.16.243.219 was causing a hard-stop whereas 204.16.243.217 was skipping one but eventually making it to a perlmonks server. But today, whichever perlmonks IP I am tracing to, it is going through .219 but is connecting.

Over the weekend at $home, I could never get a connection using my home network. But if I took my phone off wifi and tried to connect with my phone's data provider instead, it could get a response and let me read posts.

And when I try https://www.isitdownrightnow.com/perlmonks.org.html, they say it's been down "~6 days" at this point.

It looks like it's highly dependent on where you are trying to access from whether or not you'll be able to get a route to a perlmonks server.

[Tux]

I am still unable to connect from home (you have my data). Now replying from another network using ssh -X

[LanX]

IIRC there were times when one was able to directly run a complicated super search by clicking a link.

AFAIR this was deactivated for performance reasons. Now we need to additionally click the [Search] button after "preloading"

Hence having links which automatically run searches can't be a desired situation.

Especially when dealing with bots.

Update

Probably some confusion, I'm not talking about the recent outages. I didn't experience any when I started this thread, tho it seems to be drifting in that direction.

I'm interested to know if fixing named links would improve performance notably.