| View previous topic :: View next topic |
| Author |
Message |
Walter EVDO Junkie
Joined: 14 Sep 2005 Posts: 307
|
Posted: Mon Mar 12, 2007 11:36 pm Post subject: Large Upload Speed Difference between Firefox and IE 7 |
|
|
I'm hoping this post will start a good discussion. If this topic has been discussed before I missed it. But it came as such a surprise to me that I thought it deserved a new post. This is a bit long I'm afraid and first a short little background, so bear with me...
Ever since I switched to Sprint and found I had a Rev. A connection, I've been somewhat curious as to why my uploads cap out in the 200 kb/sec range and so many others seem to be limited to about the same speed while many others report 400 up to over 700 kb/sec. I think the best I had seen (until last night, more on that in a minute) was about 230 kb/sec. This even after I added an omni antenna (from 3GStore) that stabilized my signal at -67dBm and with downloads that top out at 1300 kb/sec.
Even after Sprint changed their routing between San Jose and Los Angeles that further improved performance for me, the uploads still stayed at under 230 kb/sec.
The 230 kb/sec number is based on multiple speed test sites. Specifically, the DSLReports site, Speedtest.net, and FrontierNet. The only one that reported higher uploads was the MyVOIPSpeed site, but that always looked to me like a compression artifact. So I have been ignoring it.
And since I use Firefox as my main browser, all these test were using Firefox. In the meantime, I've been chasing other issues and a long list of other things to do and really didn't spend much time on this.
But then last night, almost on a fluke, I ran IE 7 instead of Firefox and tried one of the speed test sites. And the upload speed was MUCH HIGHER THAN 230 kb/sec! I was seeing well over 400 kb/sec.
This came as a great surprise. I tried multiple times on this particular site. And it was repeatable. Went back to Firefox and the same site... back to around 200 kb/sec. Went back to IE 7 and back to much higher uploads. Tried the other sites I regularly use. They all reported higher as well! And repeatably!
Here's an example. The image below is using Firefox:
Now here's the same site a minute or so later using IE 7:
The other sites have similar results.
So first, has anyone else seen this type of behavior? Also, is anyone reporting uploads in the 400 to 700 kb/sec using Firefox when they are getting those numbers? Can anyone else who has had uploads consistently in the 200 kb/sec range and using Firefox try the same test to see if they see similar results when using IE 7?
At this time I'm not quite sure just what is going on. A while back I had done an FTP upload test to a remote site as a check against the speed test site numbers. The FTP upload speeds seemed to be in line with a 200kb/sec or so speed. But if so, then what in the world is going on with IE 7 and are these higher speeds real? And if they are real, I'd like to figure out how to get Firefox to perform the same. And if the speeds are not real, is anyone out there *really* getting 700 kb/sec upload speeds?
Right now I have far more questions than answers. OK, everyone. Discuss! _________________ Reach, Connect, Celebrate, Grow, Serve. |
|
| Back to top |
|
 |
Michael Site Admin
Joined: 13 Jan 2005 Posts: 5045 Location: Cary, IL
|
|
| Back to top |
|
 |
horsessmellfunny EVDO Junkie
Joined: 03 Feb 2006 Posts: 356 Location: Castle Rock, CO
|
Posted: Tue Mar 13, 2007 1:23 am Post subject: |
|
|
TCP/IP Settings. Look for the TCP IP optimization thread. Make the adjustments and you will see the difference in both browsers... I bleive I left an explannation in one of the other threads regarding this issue. _________________
|
|
| Back to top |
|
 |
Kovster EVDO User
Joined: 05 Dec 2006 Posts: 84
|
Posted: Tue Mar 13, 2007 5:21 am Post subject: |
|
|
I was gonna post a topic on the same thing.
It was even giving different results with IE6, its just not IE7. |
|
| Back to top |
|
 |
rickster EVDO Junkie
Joined: 27 Aug 2005 Posts: 216
|
Posted: Tue Mar 13, 2007 5:44 am Post subject: |
|
|
I'm guessing since the difference is firefox vs IE this is not an EVDO issue nor an IPStack issue. I'm thinking this is more likely a FF configuration/optimization issue. One interesting test you could run if you have access to a faster connection (like wifi coffee shop) see if you see the difference there.
There are large number of items that can be configured in FF; try "about:config" in the URL. I'm not a FF guru and you might try http://forums.mozillazine.org/ asking on one of the forums there. I run a plugin called Fasterfox 2.0 that helps optimize some of those.
Good luck |
|
| Back to top |
|
 |
Michael Site Admin
Joined: 13 Jan 2005 Posts: 5045 Location: Cary, IL
|
Posted: Tue Mar 13, 2007 5:49 am Post subject: |
|
|
| horsessmellfunny wrote: | | TCP/IP Settings. Look for the TCP IP optimization thread. Make the adjustments and you will see the difference in both browsers... I bleive I left an explannation in one of the other threads regarding this issue. |
TCP/IP Settings are global to your computer, NOT specific to applications. _________________ EVDO :: EVDO News :: EVDO Antennas :: Buy Verizon :: Buy Sprint :: EVDO Amplifier |
|
| Back to top |
|
 |
justanoldguy EVDO Fledgling
Joined: 05 Mar 2007 Posts: 14
|
Posted: Tue Mar 13, 2007 7:29 am Post subject: |
|
|
Using IE7:
Using Firefox 2.0.0.2
 |
|
| Back to top |
|
 |
jackrodgers EVDO Addict
Joined: 23 Mar 2006 Posts: 1131
|
Posted: Tue Mar 13, 2007 8:01 am Post subject: |
|
|
Using Safari on a MacBook Pro from Okeechobee FL:
Speedtest.net now shows a popup with a list of the IP addresses used and when selected shows the phone company name, etc.
I've used about 40 different IPs with a wide range of results.
You might also check to see if your browsers are using the same ip address. |
|
| Back to top |
|
 |
grant EVDO User
Joined: 21 Jan 2005 Posts: 50
|
Posted: Tue Mar 13, 2007 8:08 am Post subject: |
|
|
| jackrodgers wrote: | | You might also check to see if your browsers are using the same ip address. |
Huh? Of course, when you quit browsers and re-launch and you haven't disconnected, you will have the SAME IP. You will usually get a different IP everytime you connect. |
|
| Back to top |
|
 |
Walter EVDO Junkie
Joined: 14 Sep 2005 Posts: 307
|
Posted: Tue Mar 13, 2007 8:39 am Post subject: |
|
|
Just to clarify for everyone: when I did my tests it was the same connection. My IP address was the same. The site addresses were the same. Everything was the same. I literally had Firefox and IE 7 running at the same time on the same machine. Did a speed test on one. Copy and pasted results into my post. Did speed test on other. Copy and pasted results to post. And so far the test is completely repeatable. It will do this everytime. There is some difference between upload values on different sites. But the IE 7 upload speeds are *always* higher and usually by at least a factor of two.
One minor clarification to Michael's post about TCP/IP settings. IIRC they are global to each network device on a computer. So if you have two NICs, you can have different TCP/IP settings for each one. But he is right that the settings are not application specific.
On top of that, the TCP/IP settings for the computer I tested on aren't really the limiting factor here since it's my server that has the EVDO device and it's not the one running Firefox or IE 7. I can assure everyone that my TCP speeds between my workstations and my server are *much* faster than 200 kb/sec. And before anyone else suggests it, yes I have tested directly to one of my workstations before to rule out the server limiting me to 200 kb/sec. When I directly connected my U720 to one of my workstations, it was also getting 200 kb/sec for uploads when I used Firefox. Haven't tried IE 7 with that setup yet (I was testing something else when I tried it), but I would fully expect the same results. The upload speeds were the same whether I used DUN or Sprint's Connection Manager.
Rickster, thanks for the tip regarding about:config. I had forgotton about that feature in Firefox. I will check it. I also know about FasterFox 2.0 but IIRC, it mostly speeds things up by downloading data in the background from URL's on the page you're viewing; so if you open one of those links, it already has the data on disk. But maybe it has some other tricks as well. I will check it again.
And thanks for the test results justanoldguy. So it looks like there's nothing inherent in Firefox limiting things to 200 kb/sec. Must be something specific to my setup/configuration. I'm still curious for a test result from someone else who's Rev. A speeds seem to top out around 200 kb/sec. _________________ Reach, Connect, Celebrate, Grow, Serve. |
|
| Back to top |
|
 |
xenophon EVDO Addict
Joined: 30 Aug 2005 Posts: 1969
|
Posted: Tue Mar 13, 2007 8:57 am Post subject: |
|
|
I just check IE6 vs. FireFox 2.0 and also see significant upload increase for IE6. At least for Java-based tests.
I was in St. Louis couple weeks ago and was able to get over 2Mbps download on just Rev0 on my sister's Vista laptop (IE7). I could only get 1.5 max on my WinXP laptop with IE6 and Firefox. |
|
| Back to top |
|
 |
Walter EVDO Junkie
Joined: 14 Sep 2005 Posts: 307
|
Posted: Tue Mar 13, 2007 9:46 am Post subject: |
|
|
Interesting xenophone. So the difference holds for IE 6 as well. And you see a difference in downloads too. I haven't seen download differences. Just upload.
The DSL Reports site I use is flash based and I see the same sort of difference there; though not as pronounced. Firefox is usually just a tad under 200 kb/sec and IE 7 is usually a little bit over 400 kb/sec. I don't know if it's significant or not that the Java based one reports higher.
Somewhat interesting in all of this is the VOIP test site. It typically reports over 700 kb/sec using both Firefox and IE 7. I discounted its results in the past, but now I'm not so sure.
I think I may end up doing some more FTP tests as a cross check again. _________________ Reach, Connect, Celebrate, Grow, Serve. |
|
| Back to top |
|
 |
jackrodgers EVDO Addict
Joined: 23 Mar 2006 Posts: 1131
|
Posted: Tue Mar 13, 2007 11:57 am Post subject: |
|
|
| grant wrote: | | jackrodgers wrote: | | You might also check to see if your browsers are using the same ip address. |
Huh? Of course, when you quit browsers and re-launch and you haven't disconnected, you will have the SAME IP. You will usually get a different IP everytime you connect. |
There are times when the telco will disconnect you and reconnect and this would produce a new ip.
Also on on a Mac we can have multiple accounts active and switch between them. I haven't tried this with Sprint to see if I have to connect new for each user account, hopefully not.
I decided to test this and on my first run I had to connect anew with the second account. the two accounts were using the same ip number in the internet connect dialog but whatismyisp returned two different account numbers.
The connections remained open in both accounts between switching user accounts for one or two tries. Then on subsequent switches I lost the connectin and had to connect again. I guess Sprint caught up with my MackBook's skills.
However, the IP number in internet connect remained the same while whatismyip returned different ip addresses. |
|
| Back to top |
|
 |
horsessmellfunny EVDO Junkie
Joined: 03 Feb 2006 Posts: 356 Location: Castle Rock, CO
|
Posted: Tue Mar 13, 2007 1:41 pm Post subject: |
|
|
When I said TCP/IP Setting, I know they are adapter specific and not application specific (I am not an idiot), but that is something to check as if your TCP/IP settings are not set properly, you will see wild speeds and weird speed fluctuations.
As far as each browser goes. Check your browser cache. IE tends to use cache a lot of things. Those speed tests typically rely on downloading and uploading an image or some sort of file. Firefox only uses 50 MB of cache. Another option to check is your proxy and connection settings on both browsers. Mmake sure they arent set to "automatically get proxy information" or something of the sort.
These browsers are totally different, so they will handle things in a totally different manner. If everything is configured properly, you should be seeing the same results. Either way, you must look at all the extenuating factors involved in this technology! |
|
| Back to top |
|
 |
rickster EVDO Junkie
Joined: 27 Aug 2005 Posts: 216
|
Posted: Tue Mar 13, 2007 2:08 pm Post subject: |
|
|
| AFAIK most of those test now use Flash or Java and get raw data that will never get cached. So affects of local caching of html and images should not affect results. |
|
| Back to top |
|
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|
 Sprint Rev A USB: Compass 597
 Sprint Rev A ExpressCard: Merlin EX720
 Purchase an V740 Rev A ExpressCard
 Sprint U727
 CTR350 Router
 CradlePoint PHS300
 Purchase a MBR1000
 Purchase an EVDO Booster Antenna
 Purchase a LinkSys 3G Router
 Purchase an EVDO Amplifier
 Your Mac EVDO Experts
|