LinkBack URL
About LinkBacksSo what the hell happened to HEXUS?? Since about midday I have not been able to access it! Was it some unplanned maintenance ? Anybody else have trouble getting on?
So what the hell happened to HEXUS?? Since about midday I have not been able to access it! Was it some unplanned maintenance ? Anybody else have trouble getting on?
Somebody forgot to put 50p in the meter
Aye, twas bust it would appear. I blame Steve.
Are the forums hosted on a different locations or something? They seem to work, but not the main site..
I think that this happened a couple of weeks ago as well....
I think it was more like somebody forgot to feed a £50 note to the meter, not 50p !
Ok who stole the money box, so they could have a sandybridge upgrade??
From our upstream...Today at 12:15 we started experiencing large volumes of packet loss to the Internet from our Network.
This was caused by a huge amount of traffic heading outbound from our dedicated servers which was
saturating our links. We were unable to trace this back to the box causing the traffic as the volume was
overloading our gigabit switches and caused one of them to suffer a complete hardware failure.
I arrived at Telehouse at 13:10 and proceeded to replace the broken switch with a standby we keep
in the rack for that eventuality. Once replaced we managed to pin point which box was causing the
traffic by shutting down all traffic and enabling each port one by one.
When the box had been pinpointed we located the scripts placed on the server by the intruder and
found the box was being used to DOS (Denial Of Service) attack a user in Germany. Our box had been
sending over a gigabit per second of traffic to the target. We removed the scripts and secured the box
and blocked outbound traffic of the protocol UDP.
We have since received an email from the network administrators in Berlin, Germany who were the
target of the attack. It appears we were one of 15 ISPs used to attack one of their IP addresses.
Full service was not restored until 16:25 for most users. This is the longest outage we have suffered
since the company began. I have put together a plan to make it easier to track issues like this back
to the source box more quickly and tomorrow I will order a replacement gigabit switch to replace
the standby we put in place.
so they took a dominoes delivery boy and asked him to watch what green light blinks the fastest and unplug that when an outage occurs in the future...
Capitalization is the difference between helping your Uncle Jack
off a horse and helping your uncle jack off a horse.
DR (20-02-2011)
It's a scary thought that even upper level internet providers can fall victim to these kind of attacks
(\___/) (\___/) (\___/) (\___/) (\___/) (\___/) (\___/)
(='.'=) (='.'=) (='.'=) (='.'=) (='.'=) (='.'=) (='.'=)
(")_(") (")_(") (")_(") (")_(") (")_(") (")_(") (")_(")
This is bunny and friends. He is fed up waiting for everyone to help him out, and decided to help himself instead!
Is it not a scarier thought that an ISP is using gigabit switches that "suffer a complete hardware failure" when they overload, rather than shutting down gracefully...?
Well it depends on the way they get hammered...
Of course, something could have broken months ago on the switch, and now was the first time it was rebooted.
PHP Code:$s = new signature();
$s->sarcasm()->intellect()->font('Courier New')->display();
so their dedicated servers all have gigabit links?
not the best way to do things.
We used gigabit uplinks for our bandwidth with 100meg connections to each server, no one server can then saturate the switch, no one switch can saturate all the links and no one server needs 1Gbit. Not really rocket science.
□ΞVΞ□
That makes sense for serving data, but then you need an additional link per server to get you gigabit between them, which is of course very useful to have.
PHP Code:$s = new signature();
$s->sarcasm()->intellect()->font('Courier New')->display();
I assumed this was colo (where you have a 100Mbps incomming and you rack your own firewall, switch etc.)
I assumed these servers would be behind their own firewall with their own dedicated switch to isolate them from the rest of the rack.
□ΞVΞ□
There's a "best way"?Originally Posted by Jay
Your scenario is fine for low usage customers, but how do you handle someone who actually wants to ship some traffic then, do you tell them you can't possibly sell them more than 100Mbps?
Usually only worth it if you're buying by the quarter/half/full rack, if you're on a per-U hosting plan then you just add more cost when the provider would give you plenty of switchports included.
but no firewall?
If you have gigabit ports on your Cisco you can use rate limiting to provide them with more than 100Mbps but having them all set to 1Gbps with a 1Gbps uplink is not really very good.
□ΞVΞ□
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote
