PDA

View Full Version : Hls 10.a



rfallatt
12-10-2012, 01:10 PM
Joe,
Thought you might want to know.
The version number cannot be read at the bottom... (must be too many revisions... small price to pay for great innovation and technical support)
For some reason I get an error message noting that "I can't see if an update exist" (I am connected to the internet.)

angus40
12-10-2012, 01:26 PM
The version numbers work fine here and update search also works fine ! must be your settings for your nic .

When Hls loads for the first time it asks for permission to be allowed to pass through the firewall make sure this is so in your settings also . :)

Lionking_Tx
12-10-2012, 01:41 PM
Wow...HLS is at 10a already ? I just downloaded 9Y this morning

Btw....i just tried the update check and it told me there is a newer version available so it seems to work fine.

rfallatt
12-10-2012, 01:44 PM
OK... So Im screwed up... Just got my work computer reloaded and all my setting are jacked.

What is "nic"

Lionking_Tx
12-10-2012, 02:18 PM
nic = network interface card

A lot of people use it wrong and say "nic card" which actually means "network interface card card" :)

Dennis Cherry
12-10-2012, 03:54 PM
I went from 9E on Friday to 10A today.

budude
12-10-2012, 04:05 PM
nic = network interface card

A lot of people use it wrong and say "nic card" which actually means "network interface card card" :)


Well - - it can be Network Interface Controller as well so NIC Card is not completely wrong but all the same anyway...

Lionking_Tx
12-10-2012, 04:15 PM
Well - - it can be Network Interface Controller as well so NIC Card is not completely wrong but all the same anyway...
I know, a lot of people say it - It's just stuck in my brain because I have a co-worker that always goes nuts when a CE or customer calls in and talks about his "nic card" :)

JHinkle
12-10-2012, 04:18 PM
I have two users using Renards that are having issues.

Don't know if its HLS, computer, or controller.

Remote debugging is tough and requires breadcrumbs.

Every new breadcrumb is a new release in an attempt to identify root-cause.

We don't have a loaf yet and we don't have root-cause ... so versions will change as more breadcrumbs are put down.

Joe

kingofkya
12-10-2012, 04:54 PM
I have two users using Renards that are having issues.

Don't know if its HLS, computer, or controller.

Remote debugging is tough and requires breadcrumbs.

Every new breadcrumb is a new release in an attempt to identify root-cause.

We don't have a loaf yet and we don't have root-cause ... so versions will change as more breadcrumbs are put down.

Joe

Thanks for your dilegent work joe.

Also joe manny of thouse usb to serial adaptors have bugs. I have thrown out quite a few at work as they can get very unstable if you send a lot of data. To the point where bsod or terminal application dies

Livermore-Dad
12-10-2012, 04:55 PM
I'm running it on a ren48LSD (no issues) renard and can hook up a ren24 if you need further assistance with this.

Tory

JHinkle
12-10-2012, 05:09 PM
Thanks for your dilegent work joe.

Also joe manny of thouse usb to serial adaptors have bugs. I have thrown out quite a few at work as they can get very unstable if you send a lot of data. To the point where bsod or terminal application dies

Thanks for the comments about the USB to COM adapter.

I was of the belief that some were flaky but had no way to confirm.

I will be speak on the phone to one of the users tonight.

Interesting failure - HLS keeps sending data throughout the song - lights freeze. HLS stops transmitting data for 15+ secs during the loading of the next sequence and the lights start working again. Its like the COM pipeline is getting clogged and opens back up if no data is sent for a period of time - or the driver is getting reset as HLS open and closes the COM port after each song.

Before I point and focus on the hardware (hardware is never the issue - its always the software) - I need to make sure HLS is working as designed.

Joe

kingofkya
12-10-2012, 06:20 PM
Thanks for the comments about the USB to COM adapter.

I was of the belief that some were flaky but had no way to confirm.

I will be speak on the phone to one of the users tonight.

Interesting failure - HLS keeps sending data throughout the song - lights freeze. HLS stops transmitting data for 15+ secs during the loading of the next sequence and the lights start working again. Its like the COM pipeline is getting clogged and opens back up if no data is sent for a period of time - or the driver is getting reset as HLS open and closes the COM port after each song.

Before I point and focus on the hardware (hardware is never the issue - its always the software) - I need to make sure HLS is working as designed.

Joe

That sounds exactly what i have experanced on the problem usb adapters. Some will just stop, if the buffer fills, or do other odd things. I have also seen one keep "Receiving" for a good 10min after i unplugged it from a router. I am thinking this is happening as HLS send empty serial data when its active. (i see serial line blinking) but vixen seams to not send data when nothing is going on.


Are the only two I would trust. ( A rebrand )
http://us.iogear.com/product/GUC232A/

http://www.ftdichip.com/Products/Cables/USBRS232.htm

Even then its easier to just get a real com port.
(also any one who wants a real serial port, get a supermicro branded board almost all of them have standard IO ports + newer chipsets.)

rush44oh1
12-10-2012, 06:35 PM
Joe
I am using renards on versoin 9A with no problems only problem i have is the rain keeps taking out my gfci outlets, but i am sure there is nothing you can do for that.
One second thought, I do have the problem when trying to play second schedule for the night but that is only problem i have and is minor to me.
i am running 192 channels of renard @115200 baud but i am using true on board com ports.
We use the usb to com port adapters at work also and i have found some flaky ones also. if i can do anything to help please let me know

jrock64
12-10-2012, 06:35 PM
Maybe it is just time to draw a line in the sand for next year and say E1.31 only.

The E1.31 to DMX/Ren boards are not that expensive and rock solid reliable.

I was struggeling with all the same issues times 4 ports.
See my previous cries for help if interested.

I modified a J1Sys Ren/DM8 to work with my setup and have not had that issue since.
The Ren/DMX4's cost about as much as 4 USB to serial converters.

Just saying what others have to be thinking.
No help or comfort for those struggling this year tho.

Joel

kingofkya
12-10-2012, 06:52 PM
Joe
I am using renards on versoin 9A with no problems only problem i have is the rain keeps taking out my gfci outlets, but i am sure there is nothing you can do for that.
One second thought, I do have the problem when trying to play second schedule for the night but that is only problem i have and is minor to me.
i am running 192 channels of renard @115200 baud but i am using true on board com ports.
We use the usb to com port adapters at work also and i have found some flaky ones also. if i can do anything to help please let me know


Yeah my scheduled is not working correctly just tried it last night for the first time but i haven't got around to really looking at what the problem could be yet.
It seams to start despite the check box that says "follow scheduled or something like that." (not sure of the words) It also doesn't stop had mine set to 5:00pm - 11:55. And it was still running at 2:00am

Some one need to do a group buy for the ftdi seral ports at the same time as renardSS that would solve so manny problems.



show run
show tech

on foundry routers/switches kills most usb adaptors:)


Joe you might want to make a com stress tester, to show users it is the adapter. about 5-20min @9600 of solid data will kill the bad adaptors.(loop back test)

cnbales
12-10-2012, 06:53 PM
jrock,

I think you give up a large group of users if you go e131 only. I have really enjoyed using HLS this year and although, I am probably heading towards e131 next year I would hate to think that most wouldn't be able to use because they are Renard only. The wonderful thing about this community is the variety of users and the many different ideas that are contributed.

JHinkle
12-10-2012, 07:11 PM
That sounds exactly what i have experanced on the problem usb adapters. Some will just stop, if the buffer fills, or do other odd things. I have also seen one keep "Receiving" for a good 10min after i unplugged it from a router. I am thinking this is happening as HLS send empty serial data when its active. (i see serial line blinking) but vixen seams to not send data when nothing is going on.


Are the only two I would trust. ( A rebrand )
http://us.iogear.com/product/GUC232A/

http://www.ftdichip.com/Products/Cables/USBRS232.htm

Even then its easier to just get a real com port.
(also any one who wants a real serial port, get a supermicro branded board almost all of them have standard IO ports + newer chipsets.)

HLS does NOT send empty serial data.

Enable Debug and you will see.

HLS only sends Serial when actively running a sequence. If you pause or stop the sequence --- Serial data stops.

E131 is different as the spec has a two sec keep alive requirement.

Joe

rfallatt
12-10-2012, 07:40 PM
tell me what to buy ...
show me how to do it...
I don't care what it is...
I want to get my show running

Sent from my ADR6300 using Tapatalk 2

JHinkle
12-10-2012, 07:50 PM
tell me what to buy ...
show me how to do it...
I don't care what it is...
I want to get my show running

Sent from my ADR6300 using Tapatalk 2

What is the issue you are try to resolve - I don't remember you ever stating it.

Joe

jimbo3301
12-10-2012, 08:03 PM
I'm using 4 E681's and 5 Ren's of various flavors. I'm using RPM's E131 to DMX bridge to drive the Ren's. I'm also driving 5, 3 Channel RGB DMX Controllers from the bridge. So far, knock on wood, I've not experienced any issues or failures.

Jim

JHinkle
12-10-2012, 08:15 PM
I'm of the belief right now that the people that are having issues with Renards are using a faulty USB to COM device.

Another user is using a USB to XBee board to drive RenWs.

All worked well until his channel count got to 155 - then issues started. We are still debugging this one to make sure its the USB device.

Joe

Livermore-Dad
12-10-2012, 08:26 PM
Maybe it is just time to draw a line in the sand for next year and say E1.31 only.

The E1.31 to DMX/Ren boards are not that expensive and rock solid reliable.

I was struggeling with all the same issues times 4 ports.
See my previous cries for help if interested.

I modified a J1Sys Ren/DM8 to work with my setup and have not had that issue since.
The Ren/DMX4's cost about as much as 4 USB to serial converters.

Just saying what others have to be thinking.
No help or comfort for those struggling this year tho.

Joel

My biggest problem is really the xbee's and communication. range in my setup is pretty bad. Lots of issues last year with that. While the e1.31 E68x cards are bullet proof (just which they would have a wifi option (as xbee is not quite a standard protocol.

But I did have quite a few issues switching to xbee with glitches misfires etc.
Tory

jrock64
12-10-2012, 09:21 PM
While the e1.31 E68x cards are bullet proof (just which they would have a wifi option (as xbee is not quite a standard protocol.

Tory

I bet that that Dirk's engineers could come up with something for that blindfolded with one hand tied behind their backs.

Just need take the Ren output from one of the Bridge boards and convert it from RS485 back to RS232 before feeding it to the zigbee.

Joel

cnbales
12-10-2012, 09:28 PM
I have 4 Ren 24SS boards, a Ren 64XC, and a Ren 48LSD and my show is running great. I am using a laptop to run my show with a USB to Com device.

jrock64
12-10-2012, 10:01 PM
I have 4 Ren 24SS boards, a Ren 64XC, and a Ren 48LSD and my show is running great. I am using a laptop to run my show with a USB to Com device.

I was fine until I tried to add a fourth com port.

There seems to be a channel number thresh hold when everything just stops working.

So the question for everyone, how many com ports and channels are you successfully running.

Cnables You do know that the bridge devices let you connect your Renard devices via E1.31.
http://www.diyledexpress.com/index.php?main_page=product_info&cPath=22&products_id=157

Joel

kingofkya
12-10-2012, 10:07 PM
Do we know witch usb xbee adaptor he is useing, as the parallax one I have has been no trouble so far.

http://www.parallax.com/Store/Accessories/CommunicationRF/tabid/161/ProductID/643/List/0/Default.aspx?SortField=ProductName,ProductName

JHinkle
12-10-2012, 10:44 PM
I was fine until I tried to add a fourth com port.

There seems to be a channel number thresh hold when everything just stops working.

So the question for everyone, how many com ports and channels are you successfully running.

Cnables You do know that the bridge devices let you connect your Renard devices via E1.31.
http://www.diyledexpress.com/index.php?main_page=product_info&cPath=22&products_id=157

Joel

Joel:

Are you using USB to COM adapters?

Are the adapters powered by FTDI?

If the answer is yes to the above - I have a posible solution - but I would need someone to test it.

Joe

jrock64
12-10-2012, 11:36 PM
Understanding that my show is successfully running off of an E1.31 board.

I can do some testing in two hours at 11pm when my show is done for the night.

I have two internal ports... seemed to work without issues.

A StarTech 4 port port pci card with 16C550 UART, Driver ComSys that would always lock up within 16 seconds max.

Two USB to serial adapters. Prolific driver. One would work, but adding the second, for 4 ports total would lock up.

A genuine enttec USB DMX Pro, uses FTDI never saw any output from this device... a light would flash when it was active, never saw it.

Now when I was testing this I could look at my switch and see the fast blink of data being transmitted,
or the slow blink that meant it was in standby, waiting to do something.

When I say locked up, the E1.31 data also stopped.

Since my show is running off of an SSD as D: , I can do any testing on my C: drive and not have to worry about corrupting my working show.

Let me know what you need.

Joel

JHinkle
12-11-2012, 12:04 AM
Joel:

If you are all E131 --- no COM --- right?

If you say everything stops ... does it ever restart on its own?

I need you to answer these questions and run your show with the latest HLS version and say YES to debug.

I need to see the debug report when then freeze occurs.

Please note song being played and time of day.

Then get me the file.

Thanks.

Joe

jrock64
12-11-2012, 12:23 AM
4 renard serial runs and one DMX run All from the E1.31 card.

Current using no serial or USB of any kind. All E1.31

When I locked up, the only way to restart was to reboot the PC.

Will download HLS10C and try some testing for you.

Joel

JHinkle
12-11-2012, 12:25 AM
Joel

Your issue is totally different than the Renard COM.

Please get a debug file that holds the freeze.

Thanks.

Joe

Lionking_Tx
12-11-2012, 01:18 AM
I'm glad I changed to (almost) full E131 this year - Only "leftover" is one entech... but that will be changed for a port on one E682 next year.

jrock64
12-11-2012, 09:58 AM
Got some of my yest results last night, then the internet was down.

Hope that I remember the scenerios well enough so here goes.

There are 6 new files in my drop box.
https://www.dropbox.com/sh/xwmyv08p8ujjo9o/FAQHEvf3fx

For all tests all output was disabled except the ports being tested,
With the exception of 1 universe of E1.31 so I could see that activity.
All testing done with HLS10C
All tests were done using the output from sequencer option.
Each performed a restart of HLS(the only way to get a new log file)

Test 1 was using the two internal serial ports.
I played the song for 1 minute, stopped, paused 30 seconds, started and played 1 minute then stopped.
PASSED

Test 2 was using the two ports on the StarTech expansion card.
Played the song for 1 minute, stopped, paused 30 seconds, started, played again.
sometime during the second play cycle the output stopped.
The lights froze and e1.31 activity stopped.

Test 3 was using the two ports on the StarTech expansion card.
REBOOT
Play complete song, stop, play 1 minute, stop, pause 30 seconds, play song 1 minute, stop
PASSED

Test 4 2 internal(3&4) and 2 expansion(1&2) StarTech
Com 1&2 were configured but may not have been ticked active, not sure.
No visible output of any kind.

Test 5 2 internal(3&4) and 2 expansion(1&2) StarTech
REBOOT
No visible output of any kind.

Test 6 disabled internal(3&4) only 2 expansion 1&2 active.
No visible output of any kind.

BEDTIME

Did not have time for any USB testing.

I do not need this fixed, just providing as information to possibly help others.

Hope it helps.

Joel

JHinkle
12-11-2012, 11:19 AM
Joel:

Thank you very much.

Just a comment - the contents of the debug file test2 suggest that HLS version 10C was not used.

There was important debugging in 10C that looked for the type of activity reported in test2 and would provide additional information.

Could you please rerun test 2 again and use 10D (not released yet).

Thanks in advance.

Joe

jrock64
12-11-2012, 11:59 AM
Maybe that's why it failed, and test 3 passed.
Test 3 was a repeat of test 2 after a reboot.
Could have used the wrong drive by mistake.

So throw out test 2 and the 2 internals or the 2 expansion by themselves work OK.
Combine them or try to output 4 ports total, and they fail.
My gut says it is the amount of data it is trying to process.
Somewhere past 400 channels it can not keep up and just stops.
Lines 1,2, and 3 average 150 channels.
Line 4 is loaded with 256 channels.

Will try the 4 expansion ports by themselves
and some USB to serial testing tonight.

Joel

JHinkle
12-11-2012, 12:17 PM
Joel:

Please make sure you get the latest version on my site prior to testing.

Thanks.

Joe

jrock64
12-12-2012, 02:25 AM
There are 7 new files in my drop box.
https://www.dropbox.com/sh/xwmyv08p8ujjo9o/FAQHEvf3fx

For all tests all output was disabled except the ports being tested,
With the exception of 1 universe of E1.31 so I could see that activity.
All testing done with HLS10D
All tests were done using the output from sequencer option.

Line 1 has 125 channels
Line 2 has 150 channels
Line 3 has 155 channels
Line 4 has 255 channels

Com 1 is internal
Com 2 is internal
Com 3 is expansion StarTech
Com 4 is expansion StarTech
Com 5 is expansion StarTech
Com 6 is expansion StarTech
Com 8 is USB to serial Prolific
Com 9 is USB to Serial Prolific

Test11 Line 1,2,3 output on com 1,2,3
REBOOT
All output stopped within 6 seconds
Fail

Test 12 Line 1,2 output on com 1,2
Restart
Play full sequence, stop, play 1 minute, stop.
Passed

Test 13 Line 1,4 output on com 1,2 high channel count
No Restart
No visible output.
Fail

Test 14 Line 4 as com2 high channel count
REBOOT
Play sequence, stop, pause, start, play 1 minute, stop
No debug file captured.
Passed.

Test 15 Line 1,4 output on com 1,2 high channel count plus other port
No Restart
No visible output
Fail

Test 16 Line 1,2,3 on com 3,4,5
REBOOT
All output stopped within 6 seconds
Fail

Test 17 Line 1,2 as com 8,9 Prolific USB
REBOOT
Start, play 2 minutes, stop
Passed

Test 18 Line 1,2,3 as com 8,9,1 prolific and internal
REBOOT
Froze within 5 seconds.
Fail

Hope this helps you solve it.


Joel

JHinkle
12-12-2012, 09:47 AM
Joel:

I very much appreciate your helping.

Looking at the reports I realized your were running from the editor - not Playlist or Show.

I did not have as much debug breadcrumbs in that path.

Version 10F - now has all me debugging in both paths.

Can I ask you to do me a favor ...

Make sure you download 10F

Run tests 11 and 16 again.

The USB adapters are causing the issue - I'm trying to see to what extent the are affecting HLS.

Thanks in advance for your help.

Joe

jrock64
12-13-2012, 02:07 AM
There are 4 new files in my drop box.
https://www.dropbox.com/sh/xwmyv08p8ujjo9o/FAQHEvf3fx

For all tests all output was disabled except the ports being tested,
With the exception of 1 universe of E1.31 so I could see that activity.
All testing done with HLS10F
All tests were done using the output from sequencer option.

Line 1 has 125 channels
Line 2 has 150 channels
Line 3 has 155 channels
Line 4 has 255 channels

Com 1 is internal
Com 2 is internal
Com 3 is expansion StarTech
Com 4 is expansion StarTech
Com 5 is expansion StarTech
Com 6 is expansion StarTech
Com 8 is USB to serial Prolific
Com 9 is USB to Serial Prolific

Test21 Line 1,2,3 output on com 1,2,3 2 internal 1 expansion
REBOOT
All output stopped within 17 seconds
Fail

Test26 Line 1,2,3 output on com 3,4,5 3 expansion
REBOOT
All output stopped within 28 seconds
Fail

Test28 Line 1,2,3 output on com 8,9,1 2 usb prolific 1 internal
REBOOT
Play full sequence, stop, play 1 minute, stop.
Passed, never expected that

Test29 Line 1,2,3 output on com 8,9,5 2 usb prolific 1 expansion
REBOOT
All output stopped within 3:00 minutes, so close
Fail

That's all the testing I can do.
Time to get out and meet the public and hand out some candy canes.

Joel

JHinkle
12-13-2012, 08:35 AM
Joel - you are the MAN!!!!

Thank you very much for your help in performing testing for me.

With your help - I FOUND it!!!!

Version 10G - fixes it.

This might even address what I first thought was totally associated with USB/COM adapters also.

It ended up being a race/boundary condition I only partially addressed.

Thanks again Joel.

Joe

jrock64
12-13-2012, 04:58 PM
Confirmed. That sucka is dead.

I threw all 685 channels at it from all four ports.

Lines 1234 as ports 1234 two internal two expansion
Played 1 and a half times
PASSED

Lines 1234 as ports 3456 all 4 expansion StarTech
Played 1 and a half times
PASSED

Lines 1234 as ports 8956 2 USB 2 expansion StarTech
Played 1 and a half times
PASSED

so what exactly is a race/boundary condition in laymans terms.

This should make for a few merry Christmases.

Joel

kingofkya
12-13-2012, 05:01 PM
Confirmed. That sucka is dead.

I threw all 685 channels at it from all four ports.

Lines 1234 as ports 1234 two internal two expansion
Played 1 and a half times
PASSED

Lines 1234 as ports 3456 all 4 expansion StarTech
Played 1 and a half times
PASSED

Lines 1234 as ports 8956 2 USB 2 expansion StarTech
Played 1 and a half times
PASSED

so what exactly is a race/boundary condition in laymans terms.

This should make for a few merry Christmases.

Joel

But the bug has big eyes.
17016

JHinkle
12-13-2012, 05:47 PM
Confirmed. That sucka is dead.

I threw all 685 channels at it from all four ports.

Lines 1234 as ports 1234 two internal two expansion
Played 1 and a half times
PASSED

Lines 1234 as ports 3456 all 4 expansion StarTech
Played 1 and a half times
PASSED

Lines 1234 as ports 8956 2 USB 2 expansion StarTech
Played 1 and a half times
PASSED

so what exactly is a race/boundary condition in laymans terms.

This should make for a few merry Christmases.

Joel

A boundary condition is when normal actions must be re-evaluated when a boundary is hit.

Example: Drive you car fast as you can ..... road ends at cliff.

Figure out when the next illumination needs to be sent ... sneak up on it so calculate it to be always be 3 millisecs early. Not expecting it to be dead ZERO is hitting a boundary condition.

Race condition: You expect A to always completes before B. Every once in a blue moon - B completes before A.

In this case ... COM based communications --- it not constant timing. Renard has values where two numbers have to be substituted for 1. Depending on how many of those values occur - the time to transmit a COM packet can vary by millisecs because the number of byte transmitted can change time tic by time tic.

My issue was when just the right COM transmission timing took place and just the right query of samples of audio processed --- like the moon and the sun in alignment just right - failure.

Once I saw the alignment of sun and moon in the debug file - it was perfectly obvious - I was embarrassed I missed it in the first place.

Joe

angus40
12-13-2012, 06:54 PM
A boundary condition is when normal actions must be re-evaluated when a boundary is hit.

Example: Drive you car fast as you can ..... road ends at cliff.

Figure out when the next illumination needs to be sent ... sneak up on it so calculate it to be always be 3 millisecs early. Not expecting it to be dead ZERO is hitting a boundary condition.

Race condition: You expect A to always completes before B. Every once in a blue moon - B completes before A.

In this case ... COM based communications --- it not constant timing. Renard has values where two numbers have to be substituted for 1. Depending on how many of those values occur - the time to transmit a COM packet can vary by millisecs because the number of byte transmitted can change time tic by time tic.

My issue was when just the right COM transmission timing took place and just the right query of samples of audio processed --- like the moon and the sun in alignment just right - failure.

Once I saw the alignment of sun and moon in the debug file - it was perfectly obvious - I was embarrassed I missed it in the first place.Joe

I had laughed out loud when I read this part . Joe there is no hand book for what you are doing here :) .... you are writing the hand book as you go .

Embarrassment is the last thing you should feel .

It just amazes me at how quickly you are putting issues to rest and improving Hls.

rfallatt
12-13-2012, 07:16 PM
Yeah... Up and running!

Sent from my ADR6300 using Tapatalk 2

cnbales
12-13-2012, 07:27 PM
I was fine until I tried to add a fourth com port.

There seems to be a channel number thresh hold when everything just stops working.

So the question for everyone, how many com ports and channels are you successfully running.

Cnables You do know that the bridge devices let you connect your Renard devices via E1.31.
http://www.diyledexpress.com/index.php?main_page=product_info&cPath=22&products_id=157

Joel

Joel,

Didn't notice your post until today... Yes, I know, I have two of RPM's DMX to e1.31 bridge PCB's. I would hate to see HLS exclude people without e1.31, which, at the present, would be the majority of the folks on here. Ultimately, this is Joe's software and he can do what he wants, but I am extremely happy that Joe made it work with the Renard protocol otherwise I might not have had a show this year.

More than anything though, I am happy that your show is up and running, bringing smiles to your and your visitors faces!

Chris