Hello and Good Morning,
Are there reports of L3 having issues this morning ? Starting at about
10:10 A Pacific, I started seeing huge drops in traffic at various
sites, including San Diego, Houston, San Antonio, Charlotte, NC,
Philadelphia, etc.
Anyone seeing a similar behavior ?
Yes we are seeing Loss from Houston To LA (not NYC to LA) dropping out in Dallas
-=Tom
Seeing similar issues here. Traffic going over Level3 seems to be dying.
В 11:39 -0600 на 01.08.2011 (пн), Khurram Khan написа:
Hello and Good Morning,
Are there reports of L3 having issues this morning ? Starting at about
10:10 A Pacific, I started seeing huge drops in traffic at various
sites, including San Diego, Houston, San Antonio, Charlotte, NC,
Philadelphia, etc.
Anyone seeing a similar behavior ?
Yes, l3 in Dallas is dropping about 40% of the traffic we're sending
them.
Yes...about the same time here...maybe a few minutes later. I was seeing traces stopping at ae-63-63.ebr3.Atlanta2.Level3.net. Things seem to be moving again. I just unshut our L3 BGP session. Interestingly, while Level3 was having problems, our local Akamai cluster was apparently not working properly. images.apple.com resolves to our local Akamai cluster, but requests there were timing out.
I happen to have an L3 link out of NYC, but unfortunately I don't have a list of on-net L3 prefixes in any of the reportedly affected regions, so I'm unable to provide any data from my vantage point up here. I'm sure others are in my position as well.
Is there any sort of etiquette/BCP for reporting issues like this to the community? Something that might specify a method of providing information a little more specific than just specifying the affected region(s)? Maybe a list of a few affected hosts/prefixes/URLS/etc?
(incidentally, images.apple.com also resolves to our local Akamai cluster)
I did finally see a Level 3 network event posted about this in their portal.
Actually they list two separate ones:
A routing issue failure between Dallas, TX and Los Angeles, CA is impacting
IP services. Impacted for: 1 hour 29 minutes
A loss of connectivity to servers in Dallas, TX, Tustin, CA, and Tokyo,
Japan caused an impact to CDN services.
The second one probably explains the Akamai issues one poster mentioned.