First hop Latency

[Tommygun]

We have a 4Mb line runing IS uncapped shaped.
Our usual latency between our router and the first hop out (dslam?) usually sits at about 35ms when the line is mostly empty.
On two other occasions and right now it jumps to about 110ms, this is a nice constant ping almost like someone's just added a 70ms overhead on our usual results.
Ive called in but no luck other than to be tolsd the line is empty and everything should be ok.
After several hours it returns to the normal 35ms again.
Anyone else? Any possible reasons, DSLAM or exchange overloaded? Im sure its nothing to do with other traffic on the line.

 

[who.is.michael]

leeusig 8 & barrack street exchanges similair...

 

[ponder]

Backhaul from DSLAM to ESR probably oversubscibed/congested.

 

[MadMailMan]

Isn't IS using VPN for it's uncapped accounts. So wouldn't that mean your first hop is actually the VPN router/server at is which is actually a few physical hops away. Or am I thinking complete bollocks here?

 

[Tommygun]

Madmailman/ponder, you could be right. My next hop is 196.212.160.85 (cpt) and dns says its IS stuff...
Should I be hassling them again do you think?

; <<>> DiG 9.5.1-P2.1 <<>> -x 196.212.160.85 any
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 40988
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0

;; QUESTION SECTION:
;85.160.212.196.in-addr.arpa. IN ANY

;; AUTHORITY SECTION:
212.196.in-addr.arpa. 522 IN SOA ns1.is.co.za. support.is.co.za. 2010060900 28800 7200 4800 600

;; Query time: 68 msec
;; SERVER: 127.0.0.1#53(127.0.0.1)
;; WHEN: Wed Jun 9 14:30:06 2010
;; MSG SIZE rcvd: 101

 

[MadMailMan]

Not really. IS have no control over what happens between your router and theirs. That is all Telscum domain and they couldn't give a chimps butthole about your latencies. Sorry for you! You MAY be able to log an issue with IS so they have ammo for their meetings with Telscum over performance. Telscum should listen to IS (eventually) as they are the largest customer.

 

[MadMailMan]

Actually you could get the IS GSD to do a trace from the router outside the VPN to see where the issue is. Might need to escalate at IS before you get an actual engineer who can do this for you.

Good luck and don't let the net bits byte.

 

[Tommygun]

Its back to normal again !

 

[Gordon_R]

Its back to normal again !

That's a sure sign of congestion IMO. Any time too many people are using an exchange/DSLAM/route/ISP, the latency will 'spike' and then return to normal later.

 

[ponder]

Its back to normal again !

Everybody just went home