![]() ![]() |
Nov 30 2007, 07:03 AM
Post
#1
|
|
|
Celery ![]() Group: Members Posts: 42 Joined: 30-November 07 Member No.: 49,813 |
I'm new to The Planet, having had two servers over at Serverbeach for a while.
I'm impressed at the amount of data available in the network graphs on the Orbit portal, but don't see a key there so I'm not getting all the understanding I could be. For example: 1. I see sometimes a switch or router is "red". Is this an indication that it is down, heavily loaded, experiencing latency, or just in my critical path? 2. I see on the provider charts a two layered graph. A green shaded area and a blue line. I don't understand from looking exactly what the relationship on those two items are. Is one an average and the other a peak? 3. Is there an overview diagram of the data center networks, so that I can see the route my packets take without having to do a traceroute and draw my own? Finally, is there a downtime history log anywhere? |
|
|
|
Nov 30 2007, 10:43 AM
Post
#2
|
|
![]() SuperGeek ![]() ![]() ![]() ![]() Group: Members Posts: 2,180 Joined: 7-November 04 From: Memphis, TN Member No.: 42,049 |
Here we go:
1. I believe that the network diagram in orbit of those routers / switches is rather old and not accurate. Red would indicate that router / switch being down, however they have made some changes but did not update Orbit with those changes. Therefore it is best to not refer to that page if looking to see if something is down. The same holds true I believe to the backbone graphs in orbit I believe. For the most part they are accurate, but some of them are missing the true capacity of that backbone. 2. Blue line = Outbound traffic.....Green shaded area = Inbound traffic. 3. I believe the closest thing to a diagram of the data center networks is in the network overview page. As far as I know, there is not a actual diagram showing the network layout / topology between all the DC's. As far as a downtime log. You would need to keep track of any downtime that your server has incurred. There is a IP monitoring system in place though that will send you an email to your Orbit email address if it is unable to query your server. |
|
|
|
Dec 1 2007, 07:58 PM
Post
#3
|
|
![]() SuperGeek ![]() Group: Admin Posts: 1,242 Joined: 18-May 07 From: Dallas, Tx Member No.: 48,459 |
1.) thedude is correct. It could also be that the router was decomissioned.
2.) thedude is spot on 3.) Due to the way our network is designed, this would be difficult to do. Reason being is that we have so much redundancy that we'd pretty much have to give you the whole network map. The route can change on the fly as routers deems necessary to provide the best route. (Insert oversimplification disclaimer) -------------------- Tomy Durden
Manager - Office of Change Management |
|
|
|
Dec 1 2007, 09:08 PM
Post
#4
|
|
![]() SuperGeek ![]() ![]() ![]() ![]() Group: Members Posts: 2,180 Joined: 7-November 04 From: Memphis, TN Member No.: 42,049 |
1.) thedude is correct. It could also be that the router was decomissioned. 2.) thedude is spot on 3.) Due to the way our network is designed, this would be difficult to do. Reason being is that we have so much redundancy that we'd pretty much have to give you the whole network map. The route can change on the fly as routers deems necessary to provide the best route. (Insert oversimplification disclaimer) It would be cool though to see an entire network map. Obviously this can't be done as you said due to routes chaning constantly depending on traffic, but it would be cool to see a "generic" map of how all the DC's are interconnected. |
|
|
|
![]() ![]() |
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:
| Lo-Fi Version | Time is now: 31st July 2010 - 07:47 AM |





Nov 30 2007, 07:03 AM






