International Traffic issues on Orcon at moment?

Forums › Orcon and iServe › International Traffic issues on Orcon at moment?

jamesrt

872 posts

Ultimate Geek

Trusted

Subscriber

# 114232 12-Feb-2013 20:31

Hi,

Anyone else experiencing international traffic problems via Orcon at the moment?

More explicitly, is anyone else able to connect to "pkg.oracle.com" at the moment?

$ traceroute pkg.oracle.com
traceroute to pkg.oraclegha.com (137.254.56.21), 64 hops max, 52 byte packets
1 bob (192.168.0.1) 1.397 ms 0.897 ms 1.121 ms
2 euba.bng1.tvc.orcon.net.nz (60.234.8.5) 36.239 ms 35.896 ms 40.881 ms
3 xe-3-3-0.cre1.sky.orcon.net.nz (121.98.9.72) 34.231 ms 33.814 ms 33.691 ms
4 * * *
5 * * *
6 * * *
^C
$ traceroute www.oracle.com
traceroute to e5734.b.akamaiedge.net (23.43.145.55), 64 hops max, 52 byte packets
1 bob (192.168.0.1) 12.678 ms 1.519 ms 1.114 ms
2 euba.bng1.tvc.orcon.net.nz (60.234.8.5) 35.368 ms 35.974 ms 40.474 ms
3 xe-3-3-0.cre1.sky.orcon.net.nz (121.98.9.72) 45.501 ms 33.571 ms 34.838 ms
4 121.98.9.137 (121.98.9.137) 34.076 ms 34.536 ms 37.607 ms
5 g7-0-1-117.tkcr4.global-gateway.net.nz (203.96.66.101) 34.611 ms 35.700 ms 35.797 ms
6 ae2-4.akcr10.global-gateway.net.nz (210.55.202.53) 35.024 ms 34.787 ms 41.822 ms
7 * * *
8 * * *
9 * * *
^C
$ traceroute www.google.com
traceroute to www.google.com (74.125.237.147), 64 hops max, 52 byte packets
1 bob (192.168.0.1) 1.255 ms 1.048 ms 1.375 ms
2 euba.bng1.tvc.orcon.net.nz (60.234.8.5) 42.965 ms 34.674 ms 37.605 ms
3 xe-3-3-0.cre1.sky.orcon.net.nz (121.98.9.72) 33.418 ms 33.728 ms 33.504 ms
4 ge-0-1-5-0.cre1.nct.orcon.net.nz (121.98.9.9) 32.966 ms 34.008 ms 34.451 ms
5 121.99.12.0 (121.99.12.0) 43.709 ms 36.010 ms 36.968 ms
6 72.14.214.5 (72.14.214.5) 70.610 ms 70.112 ms 114.396 ms
7 74.125.50.133 (74.125.50.133) 67.292 ms 119.819 ms 70.705 ms
8 66.249.95.234 (66.249.95.234) 111.111 ms 65.156 ms 68.826 ms
9 72.14.237.137 (72.14.237.137) 65.514 ms 74.493 ms 73.448 ms
10 www.google.com (74.125.237.147) 61.266 ms 69.175 ms 62.759 ms

Nothing on the Network Status page apart from a message that was yesterday; and this was working OK about 6pm this evening...

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | ... | 49

1413 posts

Uber Geek

# 760367 12-Feb-2013 20:40

downforeveryoneorjustme.com/pkg.oracle.com

mercutio

1387 posts

Uber Geek

# 760409 12-Feb-2013 21:20

yah should go with openindiana :)

jamesrt

872 posts

Ultimate Geek

Trusted

Subscriber

# 760410 12-Feb-2013 21:21

mercutio: yah should go with openindiana :)

:-)

Sounddude

I fix stuff!
1789 posts

Uber Geek

Trusted

Vocus

Subscriber

# 760417 12-Feb-2013 21:35

What DNS server are you using?

Thats tracerouting to Telecoms Akamai servers, which seems odd.

jamesrt

872 posts

Ultimate Geek

Trusted

Subscriber

# 760420 12-Feb-2013 21:40

Sounddude: What DNS server are you using?

Primary DNS: 121.98.0.1
Secondary DNS: 121.98.0.2

I assume those are Orcon's; it a Genius modem so basically black-box and I've not twiddled anything in that line.

Sounddude: Thats tracerouting to Telecoms Akamai servers, which seems odd.

Pass; really can't be bothered trying to plumb the intricacies of international load balancing and caching.

And whilst Yitz's answer was a little terse, it did show "down for everyone". Typical Oracle vendor reliability.

mercutio

1387 posts

Uber Geek

# 760421 12-Feb-2013 21:51

jamesrt:
mercutio: yah should go with openindiana :)

:-)

fwiw i did try from multiple locations, and no-one seems to have a route to it. oracle issue.

mercutio

1387 posts

Uber Geek

# 760422 12-Feb-2013 21:52

Sounddude: What DNS server are you using?

Thats tracerouting to Telecoms Akamai servers, which seems odd.

Orcon's DNS gives that same IP...

Using domain server:
Name: 121.98.0.1
Address: 121.98.0.1#53
Aliases:

www.oracle.com is an alias for www.oracle.com.edgekey.net.
www.oracle.com.edgekey.net is an alias for e5734.b.akamaiedge.net.
e5734.b.akamaiedge.net has address 23.43.145.55

mercutio

1387 posts

Uber Geek

# 760424 12-Feb-2013 21:56

jamesrt:
Sounddude: What DNS server are you using?

Primary DNS: 121.98.0.1
Secondary DNS: 121.98.0.2

I assume those are Orcon's; it a Genius modem so basically black-box and I've not twiddled anything in that line.

Sounddude: Thats tracerouting to Telecoms Akamai servers, which seems odd.

Pass; really can't be bothered trying to plumb the intricacies of international load balancing and caching.

And whilst Yitz's answer was a little terse, it did show "down for everyone". Typical Oracle vendor reliability.

The fact it stops really early in every network i connected from suggests that there's no BGP route being advertised. IE, no-one knows who hosts it... where it is. Suggesting a network type problem, rather than hitting a server that's overloaded or such.

And I do realise switching to OpenIndiana is complicated if zpool/zfs has been upgraded.. but the core staff behind ZFS etc left! The creator of Java left. It's a sinking ship. And it's unlikely that there'll be much thought put into redundant software archives etc without a community spirit.

jamesrt

872 posts

Ultimate Geek

Trusted

Subscriber

# 760529 13-Feb-2013 08:25

mercutio: The fact it stops really early in every network i connected from suggests that there's no BGP route being advertised. IE, no-one knows who hosts it... where it is. Suggesting a network type problem, rather than hitting a server that's overloaded or such.

Interestingly enough, http://www.downforeveryoneorjustme.com/pkg.oracle.com still reports down; but it's now working for me via my work internet connection. And last night http://www.site24x7.com/check-website-availability.html also reported it dead; but is now saying "OK". I guess there are still intermittent network issues. Time to lean on an Oracle contact or two and see what they say...

mercutio: And I do realise switching to OpenIndiana is complicated if zpool/zfs has been upgraded.. but the core staff behind ZFS etc left! The creator of Java left. It's a sinking ship. And it's unlikely that there'll be much thought put into redundant software archives etc without a community spirit.

Well, at the end of the day, I'm at a client site, and they are a (mostly) Oracle/Solaris shop; I don't think that's going to change in a hurry. The main issue (which I've noticed especially here in NZ) with "Open<FOO>" is the lack of good "paid support" options for corporates.

Many large organisations I've worked with want to be able to phone someone at 2am if/when their <FOO> fails; and thus avoid OpenSource options which might otherwise be quite good.

This is where RedHat does well - their <FOO> is just as "Open" as anyone's, but their paid support offerings provide high "comfort" factor - this is important to non-IT-Literate Senior Managers....

mercutio

1387 posts

Uber Geek

# 760532 13-Feb-2013 08:30

jamesrt:
mercutio: The fact it stops really early in every network i connected from suggests that there's no BGP route being advertised. IE, no-one knows who hosts it... where it is. Suggesting a network type problem, rather than hitting a server that's overloaded or such.

Interestingly enough, http://www.downforeveryoneorjustme.com/pkg.oracle.com still reports down; but it's now working for me via my work internet connection. And last night http://www.site24x7.com/check-website-availability.html also reported it dead; but is now saying "OK". I guess there are still intermittent network issues. Time to lean on an Oracle contact or two and see what they say...

Seems to work with curl for me now.

mercutio: And I do realise switching to OpenIndiana is complicated if zpool/zfs has been upgraded.. but the core staff behind ZFS etc left! The creator of Java left. It's a sinking ship. And it's unlikely that there'll be much thought put into redundant software archives etc without a community spirit.

Well, at the end of the day, I'm at a client site, and they are a (mostly) Oracle/Solaris shop; I don't think that's going to change in a hurry. The main issue (which I've noticed especially here in NZ) with "Open<FOO>" is the lack of good "paid support" options for corporates.

Many large organisations I've worked with want to be able to phone someone at 2am if/when their <FOO> fails; and thus avoid OpenSource options which might otherwise be quite good.

This is where RedHat does well - their <FOO> is just as "Open" as anyone's, but their paid support offerings provide high "comfort" factor - this is important to non-IT-Literate Senior Managers....

True. But if someone can do RedHat but not Centos, or Solaris but not OpenIndiana they either have a very narrow skill set, or it's policy.

Zeon

3513 posts

Uber Geek

Trusted

# 761536 13-Feb-2013 08:40

Yes happening here to all destinations (including geekzone for a bit which routes via Orcon). I was working this morning at around 03:00 and it looks like core router 2 just dropped as I couldn't connect to a variety of stuff at the datacentre (all off difference connections to their distribution switch), came back about 20 minutes later.

Monitoring a key server peered directly at LAX Equinix and no connectivity. IPv6 does looks fine though. Yay for IPv6!

Speedtest 2019-10-14

mercutio

1387 posts

Uber Geek

# 761538 13-Feb-2013 08:44

Zeon: Yes happening here to all destinations (including geekzone for a bit which routes via Orcon). I was working this morning at around 03:00 and it looks like core router 2 just dropped as I couldn't connect to a variety of stuff at the datacentre (all off difference connections to their distribution switch), came back about 20 minutes later.

Monitoring a key server peered directly at LAX Equinix and no connectivity. IPv6 does looks fine though. Yay for IPv6!

oh that's what those alerts were about :/

it happened about a week ago too.

Sounddude

I fix stuff!
1789 posts

Uber Geek

Trusted

Vocus

Subscriber

# 761539 13-Feb-2013 08:55

Zeon: Yes happening here to all destinations (including geekzone for a bit which routes via Orcon). I was working this morning at around 03:00 and it looks like core router 2 just dropped as I couldn't connect to a variety of stuff at the datacentre (all off difference connections to their distribution switch), came back about 20 minutes later.

Monitoring a key server peered directly at LAX Equinix and no connectivity. IPv6 does looks fine though. Yay for IPv6!

Upgrade work was taking place on our Core routers at that time.

mercutio

1387 posts

Uber Geek

# 761542 13-Feb-2013 08:58

Sounddude:
Zeon: Yes happening here to all destinations (including geekzone for a bit which routes via Orcon). I was working this morning at around 03:00 and it looks like core router 2 just dropped as I couldn't connect to a variety of stuff at the datacentre (all off difference connections to their distribution switch), came back about 20 minutes later.

Monitoring a key server peered directly at LAX Equinix and no connectivity. IPv6 does looks fine though. Yay for IPv6!

Upgrade work was taking place on our Core routers at that time.

the juniper security patch?

Zeon

3513 posts

Uber Geek

Trusted

# 761544 13-Feb-2013 09:02

Sounddude:
Zeon: Yes happening here to all destinations (including geekzone for a bit which routes via Orcon). I was working this morning at around 03:00 and it looks like core router 2 just dropped as I couldn't connect to a variety of stuff at the datacentre (all off difference connections to their distribution switch), came back about 20 minutes later.

Monitoring a key server peered directly at LAX Equinix and no connectivity. IPv6 does looks fine though. Yay for IPv6!

Upgrade work was taking place on our Core routers at that time.

Yea I assumed as much and wasn't so worried.

But yea even though the notices say its fixed I'm still not seeing connectivity. a Trace route fails at core router so perhaps BGP issue for international routes then?

Speedtest 2019-10-14

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | ... | 49