One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
The ITU-T recommendation is that ?up to 150 ms mouth-to-ear delay can be tolerated by the human ear with virtually no quality loss? and in my experience, that?s held true. They are definitely going to notice the delay. From: VoiceOps <voiceops-bounces at voiceops.org> on behalf of Carlos Alvarez <caalvarez at gmail.com> Date: Wednesday, March 22, 2017 at 1:59 PM To: "voiceops at voiceops.org" <voiceops at voiceops.org> Subject: [VoiceOps] Maximum latency One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
The reality is that you can?t get away from it. A private circuit may cut it by 10% and provide more stable Jitter? but that?s it. So your client and 500,000 other agents in the Philippines and India are in the same boat. Where latency gets really nasty and there is some scope to optimize is on call transfers/conferences. Make sure there is never hairpinning through the off-shore call centre. Or that call transfers or conferences have to double up on the path. It will depend on where the media gateways for the solution reside and whether you have any optimization when multiple call legs are made through your service. Not sure if this helps? Best Regards, Ivan Kovacevic Vice President, Client Services Star Telecom | www.startelecom.ca <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> | SIP Based Services for Contact Centers | LinkedIn <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> *From:* VoiceOps [mailto:voiceops-bounces at voiceops.org] *On Behalf Of *Carlos Alvarez *Sent:* March 22, 2017 1:59 PM *To:* voiceops at voiceops.org *Subject:* [VoiceOps] Maximum latency One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality. [image: http://t.sidekickopen61.com/e1t/o/5/f18dQhb0S7ks8dDMPbW2n0x6l2B9gXrN7sKj6v4L...]
Those are some excellent points, Ivan. They do indeed transfer calls regularly between agents in various US cities, so I assume the foreign ones will also. I'll have to ask. The US agents are on MPLS to us, 2-3ms at most. Also, my satellite phone seems to have around 700ms delay, which is quite challenging. I should try synthesizing 250ms for them and let them try it. Not sure how, but assume there's some open source out there to do it. On Wed, Mar 22, 2017 at 11:08 AM, Ivan Kovacevic < ivan.kovacevic at startelecom.ca> wrote:
The reality is that you can?t get away from it. A private circuit may cut it by 10% and provide more stable Jitter? but that?s it. So your client and 500,000 other agents in the Philippines and India are in the same boat.
Where latency gets really nasty and there is some scope to optimize is on call transfers/conferences. Make sure there is never hairpinning through the off-shore call centre. Or that call transfers or conferences have to double up on the path. It will depend on where the media gateways for the solution reside and whether you have any optimization when multiple call legs are made through your service.
Not sure if this helps?
Best Regards,
Ivan Kovacevic
Vice President, Client Services
Star Telecom | www.startelecom.ca <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> | SIP Based Services for Contact Centers | LinkedIn <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...>
*From:* VoiceOps [mailto:voiceops-bounces at voiceops.org] *On Behalf Of *Carlos Alvarez *Sent:* March 22, 2017 1:59 PM *To:* voiceops at voiceops.org *Subject:* [VoiceOps] Maximum latency
One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
[image: http://t.sidekickopen61.com/e1t/o/5/f18dQhb0S7ks8dDMPbW2n0x6l2B9gXrN7sKj6v4L...]
I use a FreeBSD box as a bump-in-the-wire to test and introduce additional latency, packet loss, and jitter using ipfw ? this is a pretty good overview of how to do it: http://fjoanis.github.io/2013/08/31/Network_Simulation_FreeBSD_DummyNet/ From: VoiceOps <voiceops-bounces at voiceops.org> on behalf of Carlos Alvarez <caalvarez at gmail.com> Date: Wednesday, March 22, 2017 at 2:13 PM To: "voiceops at voiceops.org" <voiceops at voiceops.org> Subject: Re: [VoiceOps] Maximum latency Those are some excellent points, Ivan. They do indeed transfer calls regularly between agents in various US cities, so I assume the foreign ones will also. I'll have to ask. The US agents are on MPLS to us, 2-3ms at most. Also, my satellite phone seems to have around 700ms delay, which is quite challenging. I should try synthesizing 250ms for them and let them try it. Not sure how, but assume there's some open source out there to do it. On Wed, Mar 22, 2017 at 11:08 AM, Ivan Kovacevic <ivan.kovacevic at startelecom.ca<mailto:ivan.kovacevic at startelecom.ca>> wrote: The reality is that you can?t get away from it. A private circuit may cut it by 10% and provide more stable Jitter? but that?s it. So your client and 500,000 other agents in the Philippines and India are in the same boat. Where latency gets really nasty and there is some scope to optimize is on call transfers/conferences. Make sure there is never hairpinning through the off-shore call centre. Or that call transfers or conferences have to double up on the path. It will depend on where the media gateways for the solution reside and whether you have any optimization when multiple call legs are made through your service. Not sure if this helps? Best Regards, Ivan Kovacevic Vice President, Client Services Star Telecom | www.startelecom.ca<http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> | SIP Based Services for Contact Centers | LinkedIn<http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> From: VoiceOps [mailto:voiceops-bounces at voiceops.org<mailto:voiceops-bounces at voiceops.org>] On Behalf Of Carlos Alvarez Sent: March 22, 2017 1:59 PM To: voiceops at voiceops.org<mailto:voiceops at voiceops.org> Subject: [VoiceOps] Maximum latency One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality. Error! Filename not specified.
NIST used to have an impairment app for injecting latency etc. ?Been many years since I've used it. From: Matthew M. Gamble <mgamble at thoughtfire.ca> To: Carlos Alvarez <caalvarez at gmail.com>; "voiceops at voiceops.org" <voiceops at voiceops.org> Sent: Wednesday, March 22, 2017 1:27 PM Subject: Re: [VoiceOps] Maximum latency #yiv3088174526 #yiv3088174526 -- _filtered #yiv3088174526 {panose-1:2 4 5 3 5 4 6 3 2 4;} _filtered #yiv3088174526 {font-family:Calibri;panose-1:2 15 5 2 2 2 4 3 2 4;}#yiv3088174526 #yiv3088174526 p.yiv3088174526MsoNormal, #yiv3088174526 li.yiv3088174526MsoNormal, #yiv3088174526 div.yiv3088174526MsoNormal {margin:0in;margin-bottom:.0001pt;font-size:12.0pt;}#yiv3088174526 a:link, #yiv3088174526 span.yiv3088174526MsoHyperlink {color:blue;text-decoration:underline;}#yiv3088174526 a:visited, #yiv3088174526 span.yiv3088174526MsoHyperlinkFollowed {color:purple;text-decoration:underline;}#yiv3088174526 span.yiv3088174526EmailStyle17 {font-family:Calibri;color:windowtext;}#yiv3088174526 span.yiv3088174526msoIns {text-decoration:underline;color:teal;}#yiv3088174526 .yiv3088174526MsoChpDefault {font-size:10.0pt;} _filtered #yiv3088174526 {margin:1.0in 1.0in 1.0in 1.0in;}#yiv3088174526 div.yiv3088174526WordSection1 {}#yiv3088174526 I use a FreeBSD box as a bump-in-the-wire to test and introduce additional latency, packet loss, and jitter using ipfw ? this is a pretty good overview of how to do it: http://fjoanis.github.io/2013/08/31/Network_Simulation_FreeBSD_DummyNet/ ? From: VoiceOps <voiceops-bounces at voiceops.org> on behalf of Carlos Alvarez <caalvarez at gmail.com> Date: Wednesday, March 22, 2017 at 2:13 PM To: "voiceops at voiceops.org" <voiceops at voiceops.org> Subject: Re: [VoiceOps] Maximum latency ? Those are some excellent points, Ivan.? They do indeed transfer calls regularly between agents in various US cities, so I assume the foreign ones will also.? I'll have to ask. ? The US agents are on MPLS to us, 2-3ms at most. ? Also, my satellite phone seems to have around 700ms delay, which is quite challenging.? I should try synthesizing 250ms for them and let them try it.? Not sure how, but assume there's some open source out there to do it. ? ? ? On Wed, Mar 22, 2017 at 11:08 AM, Ivan Kovacevic <ivan.kovacevic at startelecom.ca> wrote: ? The reality is that you can?t get away from it. A private circuit may cut it by 10% and provide more stable Jitter? but that?s it. So your client and 500,000 other agents in the Philippines and India are in the same boat. ? Where latency gets really nasty and there is some scope to optimize is on call transfers/conferences. Make sure there is never hairpinning through the off-shore call centre. Or that call transfers or conferences have to double up on the path. It will depend on where the media gateways for the solution reside and whether you have any optimization when multiple call legs are made through your service. ? Not sure if this helps? ? Best Regards, ? Ivan Kovacevic Vice President, Client Services Star Telecom |www.startelecom.ca | SIP Based Services for Contact Centers | LinkedIn ? From: VoiceOps [mailto:voiceops-bounces at voiceops.org]On Behalf Of Carlos Alvarez Sent: March 22, 2017 1:59 PM To: voiceops at voiceops.org Subject: [VoiceOps] Maximum latency ? One of our larger customers is about to launch a new call center in Malaysia.? The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms.? We've never knowingly had a connection over 130ms.? Does anyone have experiences, good or bad, with latency approaching a quarter second?? The jitter level seems fine, so I believe they'll just have a delay but decent call quality. ? Error! Filename not specified. ? _______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops
What is it that makes people think you need a special application or tool to do these things? Here's my default gateway on my home LAN: sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=0.238 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=0.618 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=0.457 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=0.474 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=0.382 ms --- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4052ms rtt min/avg/max/mdev = 0.238/0.433/0.618/0.126 ms And here's my default gateway on my home LAN with 300 ms of latency "injected": sasha at mouse ~> sudo tc qdisc add dev enp1s0 root netem delay 300ms sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=300 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=300 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=300 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=301 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=300 ms --- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4005ms rtt min/avg/max/mdev = 300.841/300.902/301.038/0.605 ms And here's 300ms latency +/- variability of 40ms: sasha at mouse ~> sudo tc qdisc add dev enp1s0 root netem delay 300ms 40ms sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=299 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=316 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=324 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=321 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=326 ms --- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4002ms rtt min/avg/max/mdev = 299.411/317.749/326.472/9.706 ms And here's back to normal: [root at mouse ~]# tc qdisc del dev enp1s0 root netem [root at mouse ~]# ping 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=0.342 ms ^C --- 172.30.105.1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.342/0.342/0.342/0.000 ms And here's "loss insertion" of 5%: sasha at mouse ~> sudo tc qdisc add dev enp1s0 root netem loss 5% sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=0.514 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=0.501 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=0.546 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=0.279 ms --- 172.30.105.1 ping statistics --- 5 packets transmitted, 4 received, 20% packet loss, time 4056ms rtt min/avg/max/mdev = 0.279/0.460/0.546/0.105 ms sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=0.732 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=0.471 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=0.474 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=0.334 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=0.367 ms --- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4020ms rtt min/avg/max/mdev = 0.334/0.475/0.732/0.141 ms No, I'm not that brilliant. I Googled it: http://stackoverflow.com/questions/614795/simulate-delayed-and-dropped-packe... App to inject latency? You people kill me sometimes. -- Alex -- Alex Balashov | Principal | Evariste Systems LLC Tel: +1-706-510-6800 / +1-800-250-5920 (toll-free) Web: http://www.evaristesys.com/, http://www.csrpswitch.com/
Cool. Didn't realize u were running voice traffic while doing that? What was the sound quality like? Sent from my iPhone
On Mar 22, 2017, at 5:49 PM, Alex Balashov <abalashov at evaristesys.com> wrote:
What is it that makes people think you need a special application or tool to do these things?
Here's my default gateway on my home LAN:
sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=0.238 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=0.618 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=0.457 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=0.474 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=0.382 ms
--- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4052ms rtt min/avg/max/mdev = 0.238/0.433/0.618/0.126 ms
And here's my default gateway on my home LAN with 300 ms of latency "injected":
sasha at mouse ~> sudo tc qdisc add dev enp1s0 root netem delay 300ms
sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=300 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=300 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=300 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=301 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=300 ms
--- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4005ms rtt min/avg/max/mdev = 300.841/300.902/301.038/0.605 ms
And here's 300ms latency +/- variability of 40ms:
sasha at mouse ~> sudo tc qdisc add dev enp1s0 root netem delay 300ms 40ms sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=299 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=316 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=324 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=321 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=326 ms
--- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4002ms rtt min/avg/max/mdev = 299.411/317.749/326.472/9.706 ms
And here's back to normal:
[root at mouse ~]# tc qdisc del dev enp1s0 root netem [root at mouse ~]# ping 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=0.342 ms ^C --- 172.30.105.1 ping statistics --- 1 packets transmitted, 1 received, 0% packet loss, time 0ms rtt min/avg/max/mdev = 0.342/0.342/0.342/0.000 ms
And here's "loss insertion" of 5%:
sasha at mouse ~> sudo tc qdisc add dev enp1s0 root netem loss 5% sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=0.514 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=0.501 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=0.546 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=0.279 ms
--- 172.30.105.1 ping statistics --- 5 packets transmitted, 4 received, 20% packet loss, time 4056ms rtt min/avg/max/mdev = 0.279/0.460/0.546/0.105 ms
sasha at mouse ~> ping -c 5 172.30.105.1 PING 172.30.105.1 (172.30.105.1) 56(84) bytes of data. 64 bytes from 172.30.105.1: icmp_seq=1 ttl=64 time=0.732 ms 64 bytes from 172.30.105.1: icmp_seq=2 ttl=64 time=0.471 ms 64 bytes from 172.30.105.1: icmp_seq=3 ttl=64 time=0.474 ms 64 bytes from 172.30.105.1: icmp_seq=4 ttl=64 time=0.334 ms 64 bytes from 172.30.105.1: icmp_seq=5 ttl=64 time=0.367 ms
--- 172.30.105.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4020ms rtt min/avg/max/mdev = 0.334/0.475/0.732/0.141 ms
No, I'm not that brilliant. I Googled it:
http://stackoverflow.com/questions/614795/simulate-delayed-and-dropped-packe...
App to inject latency? You people kill me sometimes.
-- Alex
-- Alex Balashov | Principal | Evariste Systems LLC
Tel: +1-706-510-6800 / +1-800-250-5920 (toll-free) Web: http://www.evaristesys.com/, http://www.csrpswitch.com/ _______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops
Coil cable in a box until the desired delay is achieved. (Reference from Flash Boys by Michael Lewis) :) On Wed, Mar 22, 2017 at 2:22 PM Carlos Alvarez <caalvarez at gmail.com> wrote:
Those are some excellent points, Ivan. They do indeed transfer calls regularly between agents in various US cities, so I assume the foreign ones will also. I'll have to ask.
The US agents are on MPLS to us, 2-3ms at most.
Also, my satellite phone seems to have around 700ms delay, which is quite challenging. I should try synthesizing 250ms for them and let them try it. Not sure how, but assume there's some open source out there to do it.
On Wed, Mar 22, 2017 at 11:08 AM, Ivan Kovacevic < ivan.kovacevic at startelecom.ca> wrote:
The reality is that you can?t get away from it. A private circuit may cut it by 10% and provide more stable Jitter? but that?s it. So your client and 500,000 other agents in the Philippines and India are in the same boat.
Where latency gets really nasty and there is some scope to optimize is on call transfers/conferences. Make sure there is never hairpinning through the off-shore call centre. Or that call transfers or conferences have to double up on the path. It will depend on where the media gateways for the solution reside and whether you have any optimization when multiple call legs are made through your service.
Not sure if this helps?
Best Regards,
Ivan Kovacevic
Vice President, Client Services
Star Telecom | www.startelecom.ca <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> | SIP Based Services for Contact Centers | LinkedIn <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...>
*From:* VoiceOps [mailto:voiceops-bounces at voiceops.org] *On Behalf Of *Carlos Alvarez *Sent:* March 22, 2017 1:59 PM *To:* voiceops at voiceops.org *Subject:* [VoiceOps] Maximum latency
One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
[image: http://t.sidekickopen61.com/e1t/o/5/f18dQhb0S7ks8dDMPbW2n0x6l2B9gXrN7sKj6v4L...]
_______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops
"We're gonna need a bigger cable..." Also, Ryan, good point about routes. Their provider there says they will optimize routes, and seems to be doing so. On our side, they are very stable. On Wed, Mar 22, 2017 at 11:25 AM, Patrick Labbett <patrick.labbett at gmail.com
wrote:
Coil cable in a box until the desired delay is achieved. (Reference from Flash Boys by Michael Lewis) :)
On Wed, Mar 22, 2017 at 2:22 PM Carlos Alvarez <caalvarez at gmail.com> wrote:
Those are some excellent points, Ivan. They do indeed transfer calls regularly between agents in various US cities, so I assume the foreign ones will also. I'll have to ask.
The US agents are on MPLS to us, 2-3ms at most.
Also, my satellite phone seems to have around 700ms delay, which is quite challenging. I should try synthesizing 250ms for them and let them try it. Not sure how, but assume there's some open source out there to do it.
On Wed, Mar 22, 2017 at 11:08 AM, Ivan Kovacevic < ivan.kovacevic at startelecom.ca> wrote:
The reality is that you can?t get away from it. A private circuit may cut it by 10% and provide more stable Jitter? but that?s it. So your client and 500,000 other agents in the Philippines and India are in the same boat.
Where latency gets really nasty and there is some scope to optimize is on call transfers/conferences. Make sure there is never hairpinning through the off-shore call centre. Or that call transfers or conferences have to double up on the path. It will depend on where the media gateways for the solution reside and whether you have any optimization when multiple call legs are made through your service.
Not sure if this helps?
Best Regards,
Ivan Kovacevic
Vice President, Client Services
Star Telecom | www.startelecom.ca <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...> | SIP Based Services for Contact Centers | LinkedIn <http://t.sidekickopen61.com/e1t/c/5/f18dQhb0S7lC8dDMPbW2n0x6l2B9nMJW7t5XYg1q...>
*From:* VoiceOps [mailto:voiceops-bounces at voiceops.org] *On Behalf Of *Carlos Alvarez *Sent:* March 22, 2017 1:59 PM *To:* voiceops at voiceops.org *Subject:* [VoiceOps] Maximum latency
One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
[image: http://t.sidekickopen61.com/e1t/o/5/f18dQhb0S7ks8dDMPbW2n0x6l2B9gXrN7sKj6v4L...]
_______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops
My own experience, having lived on the other side of the world for two years while running my calls through an Atlanta-based PBX, is that high latency is seldom a problem unless it's consistent. It's variation of any kind which concerns me. It leads to jitter, audio drop-outs from retraining jitter buffers, and out-of-order frames. -- Alex Balashov | Principal | Evariste Systems LLC Tel: +1-706-510-6800 / +1-800-250-5920 (toll-free) Web: http://www.evaristesys.com/, http://www.csrpswitch.com/
On Wed, Mar 22, 2017 at 02:45:27PM -0400, Alex Balashov wrote:
My own experience, having lived on the other side of the world for two years while running my calls through an Atlanta-based PBX, is that high latency is seldom a problem unless it's consistent.
s/consistent/inconsistent/g - oops. -- Alex Balashov | Principal | Evariste Systems LLC Tel: +1-706-510-6800 / +1-800-250-5920 (toll-free) Web: http://www.evaristesys.com/, http://www.csrpswitch.com/
FYI, We ran production traffic to India over private STM1 line ( data STM1) . Latancy for Ethernet over private STM1 was 190 ms ( less then 5ms jitter ). Voice quality was just fine. On Wed, Mar 22, 2017 at 1:59 PM, Carlos Alvarez <caalvarez at gmail.com> wrote:
One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
_______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops
Have had several customers do this. No real issues other than the concerns you already voiced. You might do some monitoring of hop-count in-path so see how stable the path is. Path instability with varying latency in the path could lead to out of order frames at peak times and get you chasing your tail. On 3/22/2017 10:59 AM, Carlos Alvarez wrote:
One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
_______________________________________________ VoiceOps mailing list VoiceOps at voiceops.org https://puck.nether.net/mailman/listinfo/voiceops
We have a couple call centers in Australia and latency is consistently at around 260ms for both sites from our switch in Texas. We don't see call quality issues. Aviv On Wed, Mar 22, 2017, at 10:59 AM, Carlos Alvarez wrote:
One of our larger customers is about to launch a new call center in Malaysia. The connection there is fast, the trace looks surprisingly clean and short, but latency is consistently at 239-245ms. We've never knowingly had a connection over 130ms. Does anyone have experiences, good or bad, with latency approaching a quarter second? The jitter level seems fine, so I believe they'll just have a delay but decent call quality.
_________________________________________________
VoiceOps mailing list
VoiceOps at voiceops.org
participants (9)
-
abalashov＠evaristesys.com -
aviv＠ironsip.com -
avorlando＠yahoo.com -
caalvarez＠gmail.com -
clecny＠gmail.com -
ivan.kovacevic＠startelecom.ca -
mgamble＠thoughtfire.ca -
patrick.labbett＠gmail.com -
ryandelgrosso＠gmail.com