RTT for Clustering over IP between two countries in Europe

Unanswered Question
May 6th, 2008

The following is a extended ping test from server in England to a server in Netherlands: The first test is ran using 1500 byte packets. The second test is ran using the default packet size.

The RTT requirement is 80ms for Clustering over IP for Call Manager 6.1, and wants to make sure if they look okay. To me they look fine, but just want opinion from forum as well.

EAMS501M#ping

Protocol [ip]:

Target IP address: 172.16.10.35

Repeat count [5]: 100

Datagram size [100]: 1500

Timeout in seconds [2]:

Extended commands [n]: y

Source address or interface:

Type of service [0]: 3

Set DF bit in IP header? [no]:

Validate reply data? [no]:

Data pattern [0xABCD]:

Loose, Strict, Record, Timestamp, Verbose[none]:

Sweep range of sizes [n]:

Type escape sequence to abort.

Sending 100, 1500-byte ICMP Echos to 156.20.147.16, timeout is 2 seconds:

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Success rate is 100 percent (100/100), round-trip min/avg/max = 36/44/92 ms

EAMS501M#ping

Protocol [ip]:

Target IP address: 172.16.10.35

Repeat count [5]:

Datagram size [100]:

Timeout in seconds [2]:

Extended commands [n]: y

Source address or interface:

Type of service [0]: 3

Set DF bit in IP header? [no]:

Validate reply data? [no]:

Data pattern [0xABCD]:

Loose, Strict, Record, Timestamp, Verbose[none]:

Sweep range of sizes [n]:

Type escape sequence to abort.

Sending 5, 100-byte ICMP Echos to 156.20.147.16, timeout is 2 seconds:

!!!!!

Success rate is 100 percent (5/5), round-trip min/avg/max = 20/21/24 ms

EAMS501M#

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Jaime Valencia Tue, 05/06/2008 - 13:34

the delay test in our docs only mentions changing the TOS to use the same QoS but both of them look fine. using the default size and the 1500 MTU looks fine, seems your QoS is configured the right way to use clustering over the WAN. =)

HTH

javalenc

if this helps, please rate

abbas.ali Tue, 05/06/2008 - 13:43

Javalenc,

Thanks ! Just to let you know that we haven't deployed Call Manager yet at the customer location, but we asked to run the ping test between the two different location in two different countries. But sites are connected via MPLS. The Call Manager Publisher and a subscriber will be deployed at the hub location which connect to MPLS via 10MB link, and the other subscriber that will be clustered over IP will be at a remote location and that remote location is connected over 2MB E1 MPLS circuit. The QOS is not configured yet, and I believe the IT folks just ran the ping tests as we told them to do so to make sure that the delay is under 80ms. Don't know how busy the MPLS circuit at the time when the pings tests were ran.

Is there any recommendation we need to take into account to fully make sure we cover the 80ms RTT in full details.

Regards,

Jaime Valencia Tue, 05/06/2008 - 17:42

i'd say to perform the test at different times thru the day, if you do this at night with no one working it's logical to get a good RTT.

you would need to try testing this during normal working hours and specially during the busiest period. QoS should be able to get the voice packets the right priority once you configure it.

about the remote location with just 2MB MPLS remember 1.544 MB must be allocated for the ICCS, so there will be only a little BW left.

all this is in the SRND under the deployment models, i'm sure you already got the doc so i would advise to take all the considerations that are in there for this to work the proper way.

HTH

javalenc

if this helps, please rate

Actions

This Discussion