New 2011 MBP bad ping times

[bcrawfo2]

I have a week old 2.2 ghz MBP and have poor wireless performance.
If I ping my default gateway (my router) I get times like below.
Other computers on the same WIFI network get constant sub 1ms responses.
Can others try this? Any ideas? Again...I'm using the same WIFI signal. I've killed all possible programs. I've rebooted. Turned off bluetooth.

Thanks for any ideas.

64 bytes from 10.1.1.1: icmp_seq=283 ttl=64 time=5.676 ms
64 bytes from 10.1.1.1: icmp_seq=284 ttl=64 time=23.477 ms
64 bytes from 10.1.1.1: icmp_seq=285 ttl=64 time=3.954 ms
64 bytes from 10.1.1.1: icmp_seq=286 ttl=64 time=71.108 ms
64 bytes from 10.1.1.1: icmp_seq=287 ttl=64 time=94.576 ms
64 bytes from 10.1.1.1: icmp_seq=288 ttl=64 time=16.220 ms
64 bytes from 10.1.1.1: icmp_seq=289 ttl=64 time=3.909 ms
64 bytes from 10.1.1.1: icmp_seq=290 ttl=64 time=62.956 ms
64 bytes from 10.1.1.1: icmp_seq=291 ttl=64 time=87.071 ms
64 bytes from 10.1.1.1: icmp_seq=292 ttl=64 time=31.555 ms
64 bytes from 10.1.1.1: icmp_seq=293 ttl=64 time=38.025 ms
64 bytes from 10.1.1.1: icmp_seq=294 ttl=64 time=55.829 ms

 

[enklined]

I get similar ping times as well. I didn't notice until I switched out my router for a Netgear WNDR3700 and then pinged it. I'll try pinging the router here at work at some point to see if I get similar results.

 

[sau]

Same problem here.

I'm at work at the moment, we use an AirPort Extreme in "802.11n (802.11b/g compatible)" mode.

PING 192.168.7.1 (192.168.7.1): 56 data bytes
64 bytes from 192.168.7.1: icmp_seq=0 ttl=64 time=64.410 ms
64 bytes from 192.168.7.1: icmp_seq=1 ttl=64 time=1.330 ms
64 bytes from 192.168.7.1: icmp_seq=2 ttl=64 time=1.043 ms
64 bytes from 192.168.7.1: icmp_seq=3 ttl=64 time=143.283 ms
64 bytes from 192.168.7.1: icmp_seq=4 ttl=64 time=58.216 ms
64 bytes from 192.168.7.1: icmp_seq=5 ttl=64 time=285.570 ms
64 bytes from 192.168.7.1: icmp_seq=6 ttl=64 time=206.988 ms
64 bytes from 192.168.7.1: icmp_seq=7 ttl=64 time=2.468 ms
64 bytes from 192.168.7.1: icmp_seq=8 ttl=64 time=1.899 ms
64 bytes from 192.168.7.1: icmp_seq=9 ttl=64 time=1.907 ms
64 bytes from 192.168.7.1: icmp_seq=10 ttl=64 time=199.732 ms
64 bytes from 192.168.7.1: icmp_seq=11 ttl=64 time=5.078 ms
64 bytes from 192.168.7.1: icmp_seq=12 ttl=64 time=42.782 ms
64 bytes from 192.168.7.1: icmp_seq=13 ttl=64 time=3.899 ms
64 bytes from 192.168.7.1: icmp_seq=14 ttl=64 time=192.518 ms
64 bytes from 192.168.7.1: icmp_seq=15 ttl=64 time=1.239 ms
^C
--- 192.168.7.1 ping statistics ---
16 packets transmitted, 16 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 1.043/75.773/285.570/93.355 ms

Brand new 15" MacBook Pro. My old MacBook Pro (Late 2008) pings are fine.

 

[mike2732]

Same problem here.

I'm at work at the moment, we use an AirPort Extreme in "802.11n (802.11b/g compatible)" mode.

PING 192.168.7.1 (192.168.7.1): 56 data bytes
64 bytes from 192.168.7.1: icmp_seq=0 ttl=64 time=64.410 ms
64 bytes from 192.168.7.1: icmp_seq=1 ttl=64 time=1.330 ms
64 bytes from 192.168.7.1: icmp_seq=2 ttl=64 time=1.043 ms
64 bytes from 192.168.7.1: icmp_seq=3 ttl=64 time=143.283 ms
64 bytes from 192.168.7.1: icmp_seq=4 ttl=64 time=58.216 ms
64 bytes from 192.168.7.1: icmp_seq=5 ttl=64 time=285.570 ms
64 bytes from 192.168.7.1: icmp_seq=6 ttl=64 time=206.988 ms
64 bytes from 192.168.7.1: icmp_seq=7 ttl=64 time=2.468 ms
64 bytes from 192.168.7.1: icmp_seq=8 ttl=64 time=1.899 ms
64 bytes from 192.168.7.1: icmp_seq=9 ttl=64 time=1.907 ms
64 bytes from 192.168.7.1: icmp_seq=10 ttl=64 time=199.732 ms
64 bytes from 192.168.7.1: icmp_seq=11 ttl=64 time=5.078 ms
64 bytes from 192.168.7.1: icmp_seq=12 ttl=64 time=42.782 ms
64 bytes from 192.168.7.1: icmp_seq=13 ttl=64 time=3.899 ms
64 bytes from 192.168.7.1: icmp_seq=14 ttl=64 time=192.518 ms
64 bytes from 192.168.7.1: icmp_seq=15 ttl=64 time=1.239 ms
^C
--- 192.168.7.1 ping statistics ---
16 packets transmitted, 16 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 1.043/75.773/285.570/93.355 ms

Brand new 15" MacBook Pro. My old MacBook Pro (Late 2008) pings are fine.

Really, do a couple hundred ms really make a big difference?

I'm reading about a lot of problems with the new MBP 2011's, Was it this bad with previous generations? Should I wait to buy one?

Edit: on my windows computer I did a pingtest, i get sub 1 ms but i noticed the "bytes=32" and not 64, possibly that makes a difference?

 

[bcrawfo2]

Update...
I went to the genius desk and they saw the same thing on their 2010 (?) MBP. They booted my Mac with a recovery disk and the problem was still there.
He did some more searching and said he found an existing case that was in progress. There was no way for me to track this, so I could only take his word for it.


Mike...you can adjust the packet size on your window machine to be whatever you want with the -s option. I adjusted on my Windows test and saw the same results.
You question whether 100ms makes a difference or not. My tests were done on an access point with NOTHING plugged into it...no users...nothing. I factory reset of the router and didn't even connect it to my internet connection. So...my pings went 10 feet and took 80 milliseconds. In contrast...a healthy system...my pings can go 1500 miles in about 70 milliseconds.

 

[Tomb01]

Network performance

Ditto on the ping situation. Have a 2009 MBP, sitting right next to my new 2011 17" 2.3, and it is averaging 1.5ms pings versus the new machine getting 51ms.

Shows up in speedtest.net, too, where my old MBP gets 25Mbps down and my new one gets 5Mbps.

 

[bcrawfo2]

Hi Tom,
Thanks for your reply.
So...the reason I looked at my ping times is I had poor wireless performance to my local server. Do you have any local resources you could do speed tests too? Maybe upload/download a file from a wired device and see if your 09 MBP has different real-life performance than your 2011 MBP.

Scott

 

[iMacDragon]

I think this may actually be a power saving feature, try pinging when you've got a constant network stream going ( though not one big enough to saturate wifi ), it will probably look a lot more like what expecting.

Also speedtest is fine for me on wifi still.

 

[bcrawfo2]

Interesting...I saw the same thing. I streamed from my local webcams and the ping times got almost perfect. I stopped looking at the webcams and they went back up.

 

[sau]

I think this may actually be a power saving feature, try pinging when you've got a constant network stream going ( though not one big enough to saturate wifi ), it will probably look a lot more like what expecting.

Also speedtest is fine for me on wifi still.

I think you're right. Ping seems much better while the connection is in use.

PING 192.168.7.1 (192.168.7.1): 56 data bytes
64 bytes from 192.168.7.1: icmp_seq=0 ttl=64 time=2.853 ms
64 bytes from 192.168.7.1: icmp_seq=1 ttl=64 time=0.941 ms
64 bytes from 192.168.7.1: icmp_seq=2 ttl=64 time=1.819 ms
64 bytes from 192.168.7.1: icmp_seq=3 ttl=64 time=8.014 ms
64 bytes from 192.168.7.1: icmp_seq=4 ttl=64 time=25.637 ms
64 bytes from 192.168.7.1: icmp_seq=5 ttl=64 time=1.403 ms
64 bytes from 192.168.7.1: icmp_seq=6 ttl=64 time=2.864 ms
64 bytes from 192.168.7.1: icmp_seq=7 ttl=64 time=1.165 ms
64 bytes from 192.168.7.1: icmp_seq=8 ttl=64 time=1.666 ms
64 bytes from 192.168.7.1: icmp_seq=9 ttl=64 time=0.988 ms
64 bytes from 192.168.7.1: icmp_seq=10 ttl=64 time=1.433 ms
64 bytes from 192.168.7.1: icmp_seq=11 ttl=64 time=1.361 ms
64 bytes from 192.168.7.1: icmp_seq=12 ttl=64 time=3.738 ms
64 bytes from 192.168.7.1: icmp_seq=13 ttl=64 time=1.165 ms
64 bytes from 192.168.7.1: icmp_seq=14 ttl=64 time=0.989 ms

--- 192.168.7.1 ping statistics ---
15 packets transmitted, 15 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.941/3.736/25.637/6.110 ms

I don't have any throughput issues (it's comparable to my late '08 Macbook Pro).

 

[Ccrew]

Really, do a couple hundred ms really make a big difference?

Does if you're using a VPN.....

 

[Tomb01]

Hi Tom,
Thanks for your reply.
So...the reason I looked at my ping times is I had poor wireless performance to my local server. Do you have any local resources you could do speed tests too? Maybe upload/download a file from a wired device and see if your 09 MBP has different real-life performance than your 2011 MBP.

Scott

Using speedtest.net, the 2009 was consistently about 25mbps (Fios connection), while the 2011 never got above 5.5mbps. Didn't matter what server I picked, same result. Went to the Genius Bar, got a knowledgable guy, who put one of his store machines next to mine, and it got consistent 4ms pings to the in store router while my machine did the same widely varying times. Then he connected it to a cable, and got solid consistent .4ms pings to the same router. He indicated it was either the wireless card (apparently on a daughter card on the new machines) or the antenna assembly which is in the display assembly. Am on the road, so could not leave it with them, but will schedule something with my home store to leave it overnight so they can change out the card/antenna. Will report back if they manage to correct the problem.

 

[revenire]

Same thing here. Two week old 2011 MBP with high PINGs.

 

[finchwizard]

If you go into your System Preferences and to where the IP address is, and change the IPV6 to Disabled/Off does it improve at all?

 

[revenire]

If you go into your System Preferences and to where the IP address is, and change the IPV6 to Disabled/Off does it improve at all?

Not one bit. Had a 2010 MBP, not an issue.

 

[lurrego]

I think you're right. Ping seems much better while the connection is in use.

PING 192.168.7.1 (192.168.7.1): 56 data bytes
64 bytes from 192.168.7.1: icmp_seq=0 ttl=64 time=2.853 ms
64 bytes from 192.168.7.1: icmp_seq=1 ttl=64 time=0.941 ms
64 bytes from 192.168.7.1: icmp_seq=2 ttl=64 time=1.819 ms
64 bytes from 192.168.7.1: icmp_seq=3 ttl=64 time=8.014 ms
64 bytes from 192.168.7.1: icmp_seq=4 ttl=64 time=25.637 ms
64 bytes from 192.168.7.1: icmp_seq=5 ttl=64 time=1.403 ms
64 bytes from 192.168.7.1: icmp_seq=6 ttl=64 time=2.864 ms
64 bytes from 192.168.7.1: icmp_seq=7 ttl=64 time=1.165 ms
64 bytes from 192.168.7.1: icmp_seq=8 ttl=64 time=1.666 ms
64 bytes from 192.168.7.1: icmp_seq=9 ttl=64 time=0.988 ms
64 bytes from 192.168.7.1: icmp_seq=10 ttl=64 time=1.433 ms
64 bytes from 192.168.7.1: icmp_seq=11 ttl=64 time=1.361 ms
64 bytes from 192.168.7.1: icmp_seq=12 ttl=64 time=3.738 ms
64 bytes from 192.168.7.1: icmp_seq=13 ttl=64 time=1.165 ms
64 bytes from 192.168.7.1: icmp_seq=14 ttl=64 time=0.989 ms

--- 192.168.7.1 ping statistics ---
15 packets transmitted, 15 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.941/3.736/25.637/6.110 ms

I don't have any throughput issues (it's comparable to my late '08 Macbook Pro).

I can attest to the wireless performance improving with the network in use.

As an example: I was downloading a file that was 136mb, and it was estimated at getting done in 1hr 30mins (...i know), then I figured I'd kill the time watching a youtube video. The video was exactly one minute long. When it was done, the download was also done. Any way to make this performance constant?

 

[tbertran]

After getting a few too many dropped Airplay connections (iPhone -> AEBS -> AE) I started looking at Wifi performance and saw the same dismal ping results reported above.
I have a late 2009 iMac, a mid 2010 MBP and mid 2010 Mac mini. ALL of them (pinging each other) saw truly erratic ping times. I tried to swap my AEBS with my old Netgear wgt624 as I thought it was dying... Surely if ALL of my machines are seeing the same poor performance, it had to be it.
Nope, same results with the old Netgear.

Now, here's the kicker: reboot your machine and all is back to normal. Does it every time for me. Them it gets gradually bad again...

Right now I set my Mac mini server to reboot every night and I'm shutting down the others more often than I put them to sleep.

Now, I still haven't fixed my origina problem (iPhone playing to Airport Extreme dropping Wifi then regains it back every so often). Suggestions welcome.

 

[Robbug]

Wireless Pings to Router Widely Variable

Moderator's note: The following 10 posts have been merged in from another thread.

I was having connection issues to a MMO earlier this week and decided to ping my router to see where the issue was. I found this (noted also by several other users on the apple board too):

PING 10.0.1.0 (10.0.1.0): 56 data bytes
64 bytes from 10.0.1.3: icmp_seq=0 ttl=255 time=77.271 ms
64 bytes from 10.0.1.1: icmp_seq=0 ttl=255 time=80.713 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=1 ttl=255 time=2.632 ms
64 bytes from 10.0.1.1: icmp_seq=1 ttl=255 time=3.198 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=2 ttl=255 time=226.646 ms
64 bytes from 10.0.1.1: icmp_seq=2 ttl=255 time=227.120 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=3 ttl=255 time=148.102 ms
64 bytes from 10.0.1.1: icmp_seq=3 ttl=255 time=148.476 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=4 ttl=255 time=69.310 ms
64 bytes from 10.0.1.1: icmp_seq=4 ttl=255 time=73.042 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=5 ttl=255 time=298.074 ms
64 bytes from 10.0.1.1: icmp_seq=5 ttl=255 time=298.607 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=6 ttl=255 time=219.029 ms
64 bytes from 10.0.1.1: icmp_seq=6 ttl=255 time=219.523 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=7 ttl=255 time=140.253 ms
64 bytes from 10.0.1.1: icmp_seq=7 ttl=255 time=140.841 ms (DUP!)
64 bytes from 10.0.1.3: icmp_seq=8 ttl=255 time=61.744 ms
64 bytes from 10.0.1.1: icmp_seq=8 ttl=255 time=62.212 ms (DUP!)

Anyways - I found out that my disconnection issues were something totally different. But...I just read this article:

Thunderbolt Macbook Pro Review

On page 7 there is an interesting point on this exact "issue".

Quote:

802.11 did present the single repeatable glitch encountered during my evaluation. Wireless ping tests between MacBook Pro and a Time Capsule base station revealed widely varying latency. In some cases, response packets lagged by more than 250ms -- latency that equates to forever in LAN terms. The lag was not present with an Ethernet connection or with a Hawking HWDN2 external USB Wi-Fi adapter.

I discovered that the problem disappears when the delay between packets is reduced to 200ms. I believe that OS X is aggressively powering down the wireless radio to extend battery life (radio is a major consumer of power), but that's opinion. Apple acknowledged my report and validated my tests, but a fix was neither deemed necessary nor made available by press time. Because the issue does not affect browsing, video or audio streaming, email, or other common network tasks, it may be purely academic, but I will continue to research it.

/Quote

Personally, I haven't seen too huge a difference with my LAN performance. Occasionally, I do drop down to slow speeds but it's not too frequent. It does get annoying.

Anyways, I thought I'd toss this out on the interwebs and see what others are saying.

Thanks for reading.

Rob

 

[Nermal]

This explains the latency issues that I've been having over the past couple of weeks! I was pinging to my ISP but I never tried going directly to my AirPort. Lo and behold, fluctuating pings!

Now I need to run an Ethernet cable...

Thank you very much for bringing this to my attention and I'll let you know if I figure anything out.

 

[Derango]

This came up here a week or two ago...it was determined that, yeah, they're powering down the wireless between packets.

And like the quote says, there's pretty much no impact on real world situations due to this decision by apple. You're only going to notice in artificial ping tests and such. It's not impacting speed or latency of sustained network connections.

 

[Nermal]

Switching to Ethernet has fixed it for me, and my games are playable again!

This came up here a week or two ago...it was determined that, yeah, they're powering down the wireless between packets.

Do you have a link to that? Are there any fixes available?

 

[Robbug]

So, I just did some testing under Win 7. Definitely an OS issue. My router pings were <7ms for all packets.

I did some ping testing with http://www.pingtest.net. Under OS X I kept getting high jitter rates with no packet loss. All my grades were D's or F's

Under Win 7, out of 3 tests, I had 1 "D" with the other 2 "A's". Jitter was very minimal. It was a marketed difference.

Under real world performance, I don't notice drop outs in Win 7 but I have noticed a couple of "hiccups" under OS X - World of Warcraft.

Anyone else wanna try this to corroborate what I am seeing under two different OSs? It appears that since the wireless drivers in Win 7 are generic and the power saving features are pretty minimal.

Anyways thanks

Rob

fake edit: testing used:

terminal ping tests to my router under OS X
cmd terminal ping tests to same

http://www.pingtest.net under OS X 3 - 5 times
same under Win 7 3 - 5 times

 

[Nermal]

Under real world performance, I don't notice drop outs in Win 7 but I have noticed a couple of "hiccups" under OS X - World of Warcraft.

WoW's pretty much unplayable for me but I suspect that it's due to the relatively high latency of ~150 ms to the US in the first place. Add one of these "AirPort spikes" and it's pushing it well over 300 and "into the yellow".

You haven't listed your country but I'm assuming that you're nearer to a WoW server and therefore the spikes just put you into the 200s. It's probably therefore not quite as big an issue for you.

 

[enklined]

Not to be rude, but try searching before posting.

This has been discussed, and some insight was brought to light in a post in THIS thread.

I can confirm that bad pings only happen when the connection is idle. As soon as I start streaming a video or downloading a file, the pings drop to sub 1ms (for the most part, never exceeding 2ms).

Don't fret.

 

[Robbug]

This came up here a week or two ago...it was determined that, yeah, they're powering down the wireless between packets.

And like the quote says, there's pretty much no impact on real world situations due to this decision by apple. You're only going to notice in artificial ping tests and such. It's not impacting speed or latency of sustained network connections.

Serious question - can the wireless router determine what is artificial testing vs sustained network usage?