Version 2.2.2 and VWS: problem

[Piero]

I just installed the 2.2.2 version with VWS: everything seemed ok, but after a number of cycles that seems random, the message "station offline - last update X (after midnight) hours ago." appears. With the previous version (2.1.0), however, everything still work well.
Test page for version 2.2.2: http://www.lunavalenza.altervista.org/s ... ges-ss.htm
Page previous version (2.1.0): http://www.lunavalenza.altervista.org/ssg/gauges-ss.htm
In addition: why in the new version when the countdown starts again the first digit is always "10"?
Thank you and best regards

Piero

[mcrossley]

There is something very odd happening with your realtime data on your 2.2.2 page.

Sometimes it is OK, other times a lot of the data is either missing or wrong. See below for an example of the 'odd' data. Notice that the UTC time is 23:00 yesterday (currently 2013-3-5 09:30) and the "date" is missing, which explains the station offline message.

The temperature value "temp" has no value just a unit string - as are all the other current values, but todays hi's/lo's are there.

The realtime data on your 2.1.0 page is working fine.

Code: Select all

data: Object
BearingRangeFrom10: "359"
BearingRangeTo10: "0"
CurrentSolarMax: "0"
LastRainTipISO: "n/a"
LastRained: "n/a"
SensorContactLost: "0"
SolarRad: "0"
SolarTM: "0"
TapptempTH: "n/a"
TapptempTL: "n/a"
Tbeaufort: "0"
TdewpointTH: "00:11"
TdewpointTL: "08:17"
TheatindexTH: "00:11"
ThourlyrainTH: "n/a"
ThumTH: "07:14"
ThumTL: "00:00"
TpressTH: "09:15"
TpressTL: "03:55"
TrrateTM: "n/a"
TtempTH: "00:11"
TtempTL: "06:47"
TwchillTL: "06:13"
TwgustTM: "09:53"
UV: "0"
WindRoseData: "[0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0]"
apptemp: ""
apptempTH: "0.0"
apptempTL: "0.0"
avgbearing: "319 °"
bearing: "0 °"
bearingTM: "319"
build: " n/a"
date: ""
dateFormat: "d/m/y"
dew: "°C"
dewpointTH: "3.2°C"
dewpointTL: "0.4°C"
domwinddir: "N"
forecast: " "
heatindex: "°C"
heatindexTH: "8.3°C"
hourlyrainTH: "0.00"
hum: "0"
humTH: "91"
humTL: "83"
humidex: "-5.6"
inhum: "0"
intemp: "°C"
press: "0.0 hPa"
pressH: "1022.4 hPa"
pressL: "1021.0 hPa"
pressTH: "1022.4 hPa"
pressTL: "1021.0 hPa"
presstrendval: "0.0 hPa"
pressunit: "hPa"
rainunit: "mm"
rfall: "0.0 mm"
rrate: "0.0"
rrateTM: "0.0"
temp: "°C"
tempTH: "5.7°C"
tempTL: "1.9°C"
temptrend: " °C/last hr"
tempunit: "C"
timeUTC: "2013,03,04,23,00,00"
ver: "9"
version: "14.01"
wchill: "°C"
wchillTL: "0.2°C"
wgust: "0.0 km/h"
wgustTM: "11.3 km/h"
windTM: "4.5 km/h"
windrun: "22.1"
windunit: "km/h"
wlatest: "0.0 km/h"
wspeed: "0.0 km/h"

[Piero]

Thanks Mark,
the two pages (2.2.2 and 2.1.0) take the realtime data from the same file wflash and wflash2, then how is it possible that a page read them correctly and the other does not?
I noticed that version 2.1.0 periodically displays the message "text file downloaded corrupt .. I'm trying again," perhaps because it tries to read the file while VWS writes.
In version 2.2.2 there perhaps some different control in this sense that it could generate the error?
But the realtime data of the files wflash and wflash2 are written in some buffer after they have been read? Maybe there is some problem with the creation of this buffer?
Sorry, I'm just making assumptions, I'm not able to fully understand the code...
I do not know how to solve the problem...

[mcrossley]

With a 3 second refresh rate you are going to have lots of collisions where the file is being written to and read at the same time. That is when you will get the "text file corrupted" message, what is happening with wflash2 data is different though. The file is complete, but data is missing or wrong. It can only be the wflash2 component that is writing it incorrectly.

Where does the wflash2 get its data from - I suspect it cannot read or access it for some reason (maybe the other wflash has it locked?)

I think you need to ask those sort of questions on the VWS forum.

[Piero]

I could not tell if wflash and wflash2 are independent.
What I do not understand is why the problem with version 2.1.0 does not occur. Why do you think?
I take this opportunity to ask you another question:why when the countdown starts again the first digit is always "10"?

[mcrossley]

The 10 comes from the "downloading..." message. That has a 10 second timeout. It flashes up
"Downloading..." "10 secs"
then displays
"<blank>" "2 secs" - you have no forecast message
"<blank>" "1 sec"
"Downloading..." "10 secs"

But with such a rapid update rate you don't really get time to see them.

[Piero]

Thanks, ok.
I posted a request to Ken on wxforum, I hope to solve the problem.
Hello and thanks again

Piero

[saratogaWX]

Hi Piero,

Depending on how you have VWS doing the WeatherFlash updates (FTP v.s. Active page) can determine how long is the exposure to having an incomplete file present when the steelseriesVWSjson.php script tries to read both wflash.txt and wflash2.txt.
Note that VWS does update the two files at different times and seemingly at different intervals.

If you have a very frequent update set (something less than 10 seconds), I do suggest you change it to 10 seconds in VWS, and to use the active page method (if you're currently using FTP) -- both actions will minimize the uncertainty of getting a complete set of files when the steelseriesVWSjson.php script runs.

Best regards,
Ken

[piconut]

Hi Piero

I had the exact same problem and could never get it figured out with VWS so I started using VirtualVP and added Cumulus into the mix so I could use the data from that to feed the gauges (and I got the added windrose too!).

Good luck with your issue and let me know if you get it figured out.

Scott

[Piero]

Cumulus? Scott, you're a traitor!!! I try to resist...
I followed the advice of Ken, comments are here: http://www.wxforum.net/index.php?topic=18037.75.
Unfortunately, the problem is not solved, but version 2.1.0 still work.
I do not have enough expertise to understand what the difference is in the code in this regard.
I fear that VWS, for the realtime section, is really to be avoided, perhaps the right solution is your own!

Piero

[mcrossley]

There is nothing substantially different in the gauge code between 2.1 and 2.2 in the way it handles the realtime text file.

Can I make a suggestion? What happens if you don't run the two in parallel, that is stop/disable (whatever it is you have to do) wflash and 2.1 setup and just leave your wflash2 & 2.2.2 running? Does that improve things? Or point your 2.2.2 installation to the file created by wflash rather wflash2.

[piconut]

Cumulus? Scott, you're a traitor!!! I try to resist...
Piero

Hi Piero...

Traitor? Well actually it may not have been apparent in my post, but I'm still running VWS in addition to Cumulus. The reason I mention this is because I noticed in your list of weather software, you didn't list VirtualVP which allows you to run multiple weather software programs (VWS, Cumulus, Weather Display, etc...) on one computer by virtually splitting up the COM port to the various chosen programs. Even if you're not planning on running multiple weather software programs, it might be a nice addition to your setup as you can test other programs while relying on whatever your "chosen" program is. And if you're already running it, then never mind!

Good luck!

Here is a link to Virtual VP: http://www.softwx.com/weather/virtualvp.html

Scott

[Piero]

Mark,
the steelseriesVWSjson.php file needs two both wflash.txt and wflash2.txt files to work, so is not possible to associate a version of gauges with one only txt file.
I renamed the folder of my website in which there was version 2.1.0 and put in its place the one with the version 2.2.2. So, if I'm not mistaken, version 2.1.0 should be isolated.
However, the error appeared. That's why I was sure that there was any difference in the code of the two versions.
I understand correctly your advice or did you mean something else?

[Piero]

Ok Scott, I forgive you! I was kidding.
I know Virtual VP, anyway thanks for the advice! I tried to use it in order to be able to use weather software concurrently on different PCs connected to the same LAN as the IP datalogger. However, the virtual IPs of virtualVP do not work, so I uninstalled it. In fact I'm trying to figure out if someone continues to look after virtualVP but apparently not.
However yours is a good idea, I was just trying not to run too much software on the same PC to not overload it, but if I can not solve these problems I will do surely like you!

Piero

[piconut]

Ok Scott, I forgive you! I was kidding.
I know Virtual VP, anyway thanks for the advice! I tried to use it in order to be able to use weather software concurrently on different PCs connected to the same LAN as the IP datalogger. However, the virtual IPs of virtualVP do not work, so I uninstalled it. In fact I'm trying to figure out if someone continues to look after virtualVP but apparently not.
However yours is a good idea, I was just trying not to run too much software on the same PC to not overload it, but if I can not solve these problems I will do surely like you!

Piero

Hey Piero...I knew you were kidding!

In regards to not running too much software on one computer, don't worry unless your computer is older than mine.

Here are my computer specs:
Dell Latitude laptop C600 (14 years old)
CPU Pentium III, 1 Ghz
RAM 1 GB
Drive: 40 GB
OS: XP

Here is what I'm running (not just installed but always running):
Virtual VP
Start Watch
VWS & Weatherflash
Cumulus
ImageSalsa
Movie Salsa & Scheduler
VWSaprs
Dimension 4 (time setter)
MS Security Essentials
Filezilla
MS Remote Desktop
Logitech webcam
Freemeter (to watch CPU and resources)
various other programs and backup software

My CPU runs between 30-50% unless it is compressing the time-lapse movie and then it maxes out for about 10 minutes. So it doesn't take much processing power to run a number of programs. One thing that is clearly different with your setup is I'm not using an IP data logger so Virtual VP is just creating additional COM ports.

Granted, to get all of this running smoothly, I tweaked the Services that run and kept checking to see how much CPU was being utilized by each program and tweaking what I could there. Also, I am planning to upgrade to a newer, old computer in the near future so that I can run more webcams and add some more "stuff"!

Good luck with your setup

Scott in Austin Texas