Time for a Truck Roll...when it gets this slow, your downloads take a hit also. Recently, upload speed flat-lined around 30 Kbps, but downloads were still great. At 12 Kbps, videos buffer more.
You can only do one task at a time, but if multiple pages are loaded, it tends to freeze when these pages are sending/receiving stuff; so, you have to view only one URL without multiple pages up/running.
In two days and eleven hours, there has only been one Loss of Sync Count; either TRIA is defective, or Exede's side has an issue. I would have thought there would be more sync counts lost. I suspect TRIA has fizzled.
If you reboot SB2, it may take around 1/2 hour to function...get online.
Update - Since Tuesday, the Tx RF Power is running about 4 dBm lower, at around 30.5dBm now.
Primetime: 6:30 PM CST
Download :: 2.3 Mbps Upload :: 57 Kbps
But, early morning today, 6:53 am, upload hit 1.27 Mbps; download 9.24 Mbps. Loss of Sync Count is essentially stable; fit for duty.
I did some zip code checking for this beam, and Exede's site indicates no new installs.
Wildblue legacy users on this beam had their speeds increased in past month or so, so this takes a hit on Exede 5 users' speeds, with slower speeds than before for Exede 5 users. I doubt if 20 Mbps download will be seen again on this beam.
Hence, this pattern suggests ViaSat has changed modulation/coding which lowers Tx RF Power observed, that there is nothing wrong with TRIA, and that uplink/downlink speeds have taken a hit, due to Wildblue Legacy users getting higher speeds.
I haven't been to any video sites yet, but uplink speed needs to be pinch higher. 60 Kbps uplink is just enough for a single computer user.
LNFZ, about 1 AM, seemed to have decent/good speeds.
It would be nice if these network engineers get it figured out. Maybe the folks over at Los Alamos National Laboratory could lend them a hand, those that built the High Performance Parallel Interface. It's all about configurations and timing...Hello...
Around 11 PM CST, and I fricking know there ain't that many folks online on WB-1-163
Note - I have hooked notebook direct to SB2, and it makes no difference.
Can anyone explain what's going on? Network configuration issue, or is SB2/TRIA defective? And yes, I even have stall outs...I used "Re Preview" in this post, and I had to wait/wait for page to be reloaded.
All the posts you made with slow upload speeds show small file sizes were used for the test (256KB or less), and all the faster results shown used 1MB file size.
To be consistent from test to test try to always use a large file size for your upload tests...like 1MB or larger..
We all experience the "speed bursts" caused by the traffic management scheme used by WB/Exede...a larger test file will tend to allow these "bursts" to give better overall results. -- Exede12, ViaSat-1, beam 342, Albuquerque Gateway, Denver AcceleNet servers
How long has this been going on for now? Been a while.
Tech calls back and suggests to call this toll-free number. CSR attempts to suggest she's a part of my local coop, so I asked her who is Kenny...FAIL, she blew the test. Kenny is the main service tech.
So, the tech looked at my stats, which I can see also, and refers me to this toll free number, even though I explained this matter has been going on since November. And CSR attempts to go thru the "drill," and I say Good-Bye.
I've used a desktop and notebook, with and without router, to verify its on ViaSat's end. Today, I'm going to do a fresh Vista install on an extra computer, and do one more set of speed-tests.
SB2 modem's middle two lights, when visiting a larger webpage, like this,
will keep on "dancing" on/on, and it takes some 20 seconds to load this page. And after surfing for awhile, these two lights will really get going, and nothing happens...wait/wait/wait...so, I power cycle SB2 modem, and then wait about 20 minutes for it to re-connect to network.
Based upon SB2 modem's activity (light dancing), I suspect uplink waveform is screwing the pooch...but, does ViaSat have a means to check how many resends are happening? If I had this data, I might know more. Ping tests all come back good...which makes me wonder if problem is defective SB2 modem, or its firmware.
It's clear to me, that SB2 diagnostic info is worthless in this matter.
With ping tests coming back good, can I conclude this is a SB2 issue, or NETWORK issue...I have no idea how that Black Box (SB2) works.
Using AMD 9650 Quad-Core (HP) using Linux Mint and Vista, along with fresh install of Linux Mint-13 on that computer, my conclusion now points to Exede's Network Management Policy is either mis-configured, or other Network issues exist.
Noon hour test conditons - First, do a testmy, then Youtube video, then back to Testmy.
When this Network Management equipment kicks in, especially with slower uplink speeds, it ends up screwing the pooch for obtaining web pages....and then those two middle lights go blink-bonkers as more/more surfing persists.
I sincerely think this is a Network Management problem that due to its clamping down, this causes other SB2 issues.
For Dankberg, who never appears at a conference without quoting the latest Cisco Systems Internet traffic report, this was all too predictable. People want to make the broadband story complicated, but its really very simple, he said in the interview. Its about bandwidth, bandwidth and bandwidth.
Hello Network Engineers, your network settings on Cisco's traffic equipment are screwing the pooch, imho.
PS: ViaSat (Exede) sent an email after I "hung up:"
"Your opinion is very important to us!"
Well, see the above, your Cisco settings are screwing the pooch...and remember,
"its really very simple, he said in the interview. Its about bandwidth, bandwidth and bandwidth.
I think you're seeing the results of WB increasing the speeds for all the old Legacy customers that share bandwidth with you on your spot-beam.
I remember when I was on the Value Pak that when the beams started to over-sell that WB took bandwidth from the upload side to allow more customers to get "better" download speeds..
My upload speed dropped to 100kbs and stayed there.
Don't have any data or knowledge sources to support that but it seems your slow upload speeds began to happen soon after they announced faster speeds for all existing Legacy customers.
For geographic areas like yours, where ViaSat-1 beams are not available, you may be stuck with this unless WB/Exede adjust their bandwidth management software. -- Exede12, ViaSat-1, beam 342, Albuquerque Gateway, Denver AcceleNet servers
Yes, effective November 19, 2012, our legacy WildBlue plans speeds were increased by more than double their original speeds. No other changes were made to plans at this time.
Exede's Beam 62 (eastern Washington) and Beam 163 (mine) are both closed for new customers. Beam 62 users experience more painfully wait/wait/wait loading, stall outs.
Web pages with lots of graphical elements do not allways load these elements, even here at BBR; if you know they are missing, you have to reload the page; that video site URL, for instance, when I loaded it tonight omitted the "center" section, and took over 20 seconds to load.
Exede's network engineers need to be moved to eastern Oregon and to say York, Ne. They need to experience the dynamic effects of SB2 modem shooting craps under low uplink speeds.
All they have to do is setup committed information rate (CIR) to around 23 Kbps on their Exede 12 system to get a byte on reality.
They need to experience the "joys" of watching in lower left hand corner of their web browser each element being loaded...see picture.
Yes, when either the upload or download speed is below about 40 kb/s, one gets partial page loads and timeouts. This was happening to me during busy times between November 2009 and Spring 2012 (insufficient servers in the gateway and the Network Management Policy). In your case something else is wrong because your download speed is good. It may very well be related to increasing speeds for legacy Wildblue users. It is possible your gateway is instructing your modem to transmit at low power because the gateway is having problems with higher upload (from the customer's perspective) bit rates. -- Wildblue Value Pack, beam 31, Riverside gateway
There is no requirement that a TRIA must die by creating packet loss. Packet loss is one clue.
I suggest you monitor your Tx IF power and Tx RF power for more anomalies like transmitting at .263 W. Was that caused by the gateway instructing your modem to output 575 nW (Tx IF power) or your gain dropping low? -- Wildblue Value Pack, beam 31, Riverside gateway
Rx SNR: 11.2 dB Rx Power: -33.4 dBm 24 hours: Loss of Sync Count 3
To keep on sending the bytes to user, a packet or two would be uplinked at some time interval.
possible your gateway is instructing your modem to transmit at low power...
Adaptive Coding/Modulation - Each SB2 modem can change which format is being used. Lower bit rate coding/modulation does not require as much Tx power. When rain happens in your area, each SB2 modem will "change gears" and use a lower bit rate coding/modulation.
On Exede 5 beams, using a higher bit rate than what legacy users have is not needed. But, in time, there may be commercial 1.2m antennas, and they could use higher bit rates. Each user is given a slice of time, so if you are using a higher bit rate coding format, you can uplink more bits per unit of time.
caused by the gateway instructing your modem to output
All users on a beam must have about same power level at satellite; gateway will turn down/up each individual's Tx output power.
Those in the center, less power is needed.
In early days of Wildblue, there were whole beam resets, and that beam would be down, until it reset itself. This was due to users in center of beam who had short cable runs. Since then, they widen legacy modems to have wider parameters for controlling TRIA power output.
They called it something like SMTS resets...long ago, but I think that was what they called it.
Yes, with other factors held constant, a higher power for the carrier wave means a faster bit rate and a lower power, means a slower bit rate. I am trying to determine whether the gateway is instructing your modem to transmit at a low power and bit rate or your TRIA is malfunctioning.
Think about it. If the gateway tells your modem to transmit 15 µW, your gain drops to 10,000 and your TRIA transmits 150 mW, then your TRIA is a screwy little rabbit. -- Wildblue Value Pack, beam 31, Riverside gateway
gateway is instructing your modem to transmit at low power because
Considering it takes some 20 minutes or so for my SB2 to "connect" after a power cycle, and assuming, as joebob42 suggested, that apparently SMTS determines an user's coding/modulation used for their SB2's uplink, then when an user's connect time takes longer on a clear day, it may be that user's SB2 unit is pre-programmed to use highest bit rate first, and then it works itself downward in coding/modulation formats until SMTS takes control of it.
If true, I'm wondering how service techs or even NRTC can evaluate which coding/modulation format is being used for coming to a judgment that a TRIA is defective. For instance, NRTC claimed in May of 2012:
...NRTC members have returned only 82 SB2 units after 90 days into shipping and so far 20 of those have been found to have problems.
Did they contact ViaSat to see SMTS data, if available, or did they just install these TRIAs for testing, and said, "see these 62 TRIAs are fit for duty?"
If these "good" 62 TRIAs were using a lower bit rate, then they were defective, and NRTC screwed the pooch. And of course, consumers who received these defective TRIAs experienced slower uplink speeds, page waits, and stall-outs.
If all true above, then SB2's diagnostics ought to show coding/modulation info, and service techs need to be educated on this topic.
Have you confused SB2 modems that were bad with Tria's?
Good point....here's the terms they used in that article:
SurfBeam 2 (SB 2) equipment
SB 2 units
Keep in mind the author's abstraction level might suggest a clueless management person. Apparently, members of NRTC send equipment back to them, and its rather doubtful a single TRIA has not failed, statistically speaking.
In any event, my point is still valid if those conditions exist. If service tech is clueless about which coding/modulation is being used, and if TRIA can still operate at lower output power, then how can a service tech assess the situation?
Question remains...is my TRIA shooting craps, or has Exede reduced my uplink speed to a crawl during primetime.