connection lost Network error: Display computer : Connection lost/No response/Dissapeared

[Romanof]

Hello everyone,

We are having issues with watchout right now and we start arriving at the edge of our knowledge…

Basically, we are getting brief random “connection lost”, “unknown commands”,"No response" or also "dissapeared" errors, that forces us to go ONFFLINE/ONLINE during the day like this picture.

It happens randomly, twice or three a day on different display each time.

 The different avenues we look(ed) at :

• 2 separate network 1Gb/10Gb
  • We got rid of the 10GB and kept only the other one, the issue continues

• VNC
  • We are using Tight VNC on producers to calibrate over Wifi

  • We know that the community encourages not to use VNC service other that Watchout but use really often this method with others kit without any problem. So i'm pretty sure the problem don't com from that.

For the moment i run the show with a producer backup with exactly the same spec and there is no issue since 3 days. So i'm wondering what's happen with my producer main.

I know that this problem is not necessarily due to a network problem and the error messages can be misleading so i try to looking inside the .log file but i don't understand what i see, so may be there is a clue inside this message you can help me to understand.

 

Attached the WATCHMAKER.log file

 

 

The differents messages seem's not coming from this issue because it's not at the same time but may be i'm wrong.

My next step to try to fixe that is to flash the OS.

 

Thank you very much at all the community

 

Thank's

WATCHMAKER_00003.log

[jfk]

1 hour ago, Romanof said:

Hello everyone,

We are having issues with watchout right now and we start arriving at the edge of our knowledge…

 

 

Basically, we are getting brief random “connection lost”, “unknown commands”,"No response" or also "dissapeared" errors, that forces us to go ONFFLINE/ONLINE during the day like this picture.

 

 

It happens randomly, twice or three a day on different display each time.

 

 The different avenues we look(ed) at :

 

• 2 separate network 1Gb/10Gb

   

  • We got rid of the 10GB and kept only the other one, the issue continues

     

• VNC

   

  • We are using Tight VNC on producers to calibrate over Wifi

     

  • We know that the community encourages not to use VNC service other that Watchout but use really often this method with others kit without any problem. So i'm pretty sure the problem don't com from that.

     

For the moment i run the show with a producer backup with exactly the same spec and there is no issue since 3 days. So i'm wondering what's happen with my producer main.

I know that this problem is not necessarily due to a network problem and the error messages can be misleading so i try to looking inside the .log file but i don't understand what i see, so may be there is a clue inside this message you can help me to understand.

 

Attached the WATCHMAKER.log file

 

 

The differents messages seem's not coming from this issue because it's not at the same time but may be i'm wrong.

My next step to try to fixe that is to flash the OS.

 

Thank you very much at all the community

 

Thank's

 

 

WATCHMAKER_00003.log 9.4 kB · 0 downloads

First, i dispute your dismissal of VNC. I will bet no other product you have run  uses the archaic DirectX 3D multi-output rendering method used by WATCHOUT. Vnc and rendering are interrelated. Furthermore, WATCHOUT uses DirectX 9, which almost no one else still uses. Any observations with other product likely mean nothing as it relates to WATCHOUT.

Just the same, it is probably something else. My best guess is you are seeing watchpoint crashes with watchdog reset. The most common cause of watchpoint crash is incorrectly encoded mpeg (AVC / mpeg 4 / h.264). unless you are expert at codec settings minutiae, avoid mpeg (AVC / mpeg 4 / h.264). If settings are wrong, mpeg movies will play, but become time bombs, eventually crashing the system. Default encoder settings for mpeg will be the wrong settings. 'Best' encoder settings for mpeg will be wrong (and best is a misnomer, multiple choices can provide the same quality depending on other settings).

[Romanof]

Hey jfk,

 

First, thank you for your answer,

Concerning the part of the codec used, we use only HAP and some HAP alpha files, so i guess the problem don't come from this part, right ?

Concerning the VNC part you right, i will try to delete it and use the vnc embeded but why that's work well on my producer backup wich have also tightvnc.

 

For the watchdog log, i see nothing inside ( on the producer side ), do i need to look inside the display side log ?

 

Do you think is possible to have something wrong in my encoding show which can create a conflict on some computer ? like play severals aux timelines in same times, severals trigger, etc....

 

[jfk]

11 minutes ago, Romanof said:

Hey jfk,

First, thank you for your answer,

Concerning the part of the codec used, we use only HAP and some HAP alpha files, so i guess the problem don't come from this part, right ?

Agreed.

11 minutes ago, Romanof said:

Concerning the VNC part you right, i will try to delete it and use the vnc embeded but why that's work well on my producer backup wich have also tightvnc.

You can try. My experience has been that uninstalling does not always clear all remnants. Hopefully you archived your tuned Windows install prior to adding VNC. If not, you may consider starting the tuning process from scratch and archiving accordingly

11 minutes ago, Romanof said:

 

For the watchdog log, i see nothing inside ( on the producer side ), do i need to look inside the display side log ?

Yes. That is the only place you may find clues to why watchpoint is failing.

11 minutes ago, Romanof said:

Do you think is possible to have something wrong in my encoding show which can create a conflict on some computer ? like play severals aux timelines in same times, severals trigger, etc....

highly unlikely

Something is taking watchpoint down. So many potential reasons. Things as basic as the OS indexing a drive in the background, checking for updates in the background, etc.

A good way to determine if it is OS tuning or media would be to attempt to duplicate on a WATCHPAX.

 

[Romanof]

3 minutes ago, jfk said:

A good way to determine if it is OS tuning or media would be to attempt to duplicate on a WATCHPAX.

You mean, used a WATCHPAX as master in cluster mode ?

[jfk]

4 minutes ago, Romanof said:

You mean, used a WATCHPAX as master in cluster mode ?

To be clear, I do not suspect any network issues. Identify the watchpoint station that is crashing in your current scenario and substitute the WATCHPAX in that position. Change nothing else. The idea is to attempt to duplicate the crashing on the substituted WATCHPAX in the same manner as you are currently observing.

[Romanof]

16 hours ago, jfk said:

To be clear, I do not suspect any network issues. Identify the watchpoint station that is crashing in your current scenario and substitute the WATCHPAX in that position. Change nothing else. The idea is to attempt to duplicate the crashing on the substituted WATCHPAX in the same manner as you are currently observing.

The problem in my actual scenario is :

• The problem appears not on the same display(watchpoint); it's a different Display(watchpoint), each time.
• There is no problem when I use my backup producer (watchmaker) yet this one has exactly the same specifications and tuning as the main one.

So I guess in this scenario, it's unlikely that the problem comes from the Displays but rather from the Producer ?

 

Do you think that the Windows clock can affect Watchout? Because the clock is not set correctly on both producers (main and backup), so I wonder if this could have an influence.

 

My last option is that my main Producer's network adapter is deffected or something is not doing the job right on the hardware side but yet everything seems to work fine, so i'm not sure of that.

[RBeddig]

I the show works fine using production computer A and shows errors using production computer B, the gear to investigate is the production computer.

Are you using the same network cable and the same port of your switch when you swap the production computers? If not, do so to rule out faulty cables or broken ports on the switch.

To analyze what's going on on the network, it could help to use Wireshark on the production computer. It will write millions of events in it's log list but you can easily filter out parts once you compare the time of the network loss.

[Steve Farris]

This may or may not be relevant, but I recently had a similar situation.  After looking at the logs on the display, it turns out the license key was disappearing at random intervals.  So something caused the license to disappear from Watchpoint, which crashed Watchpoint and then Watchdog restarted it.  After rebooting the display, I had no more problems. My guess is something about the order that the key driver loaded caused an issue.

BTW, I also have TightVNC on the display.  Again, may or may not be relevant.

[wiesemann]

Sometimes USB ports go weird or defective. For example I have a surface pro dock, that cannot safely be used für USB-drives.. So I wouldnt use that. The surfaces own USB is good.

[Steve Farris]

I had the license key in a USB 3.0 port, once I moved it to an older 2.0 port all was good.

[Romanof]

Hello everyone,

Thank you for all your advise. I continu my investigation with the helps of all your comments and let you know when it's fixe.

For the moment, my problem is still here but i currently looking inside de media contents if there is no one corrupted.