Jump to content
Dataton Forum
caleidocane

Watchout startUp problem of authentication in cluster machines

Recommended Posts

Hi to everyone,

I'm stuck with a problem on a project. Due to windows problem I have re installed windows 7 on two server machines (old watchout servers with updated graphic cards FireProW5100). Before the reinstallation some bios problems but when started was ok.

This two server are in a network with all the museum and has 10.3.6.xx ip's. I've set the cluster name instead of ip's address. I've updated watch out at 6.4 release and followed the tweeting guide.

On the startup (the two display server are without production) the script is always the simplest auth->delay->load->wait->run the display that as the script start, then after variable time (20 sec circa) start the second display paused.  In this museum a device launch some string to trigger some aux timeline. If I send the string (auth->run) start perfectly in both the display without problem and then return in the main loop perfectly. Why? I'm missing something? I've tried to send directly the tcp strings to have some debug, and apparently when I send the command load "myproj" and the ask the status with get status the display response is an empty "" project loaded.

Share this post


Link to post
Share on other sites
On 6/5/2019 at 6:45 PM, caleidocane said:

... This two server are in a network with all the museum and has 10.3.6.xx ip's. I've set the cluster name instead of ip's address. I've updated watch out at 6.4 release and followed the tweeting guide.

On the startup (the two display server are without production) the script is always the simplest auth->delay->load->wait->run the display that as the script start, then after variable time (20 sec circa) start the second display paused.  In this museum a device launch some string to trigger some aux timeline. If I send the string (auth->run) start perfectly in both the display without problem and then return in the main loop perfectly. Why? I'm missing something? I've tried to send directly the tcp strings to have some debug, and apparently when I send the command load "myproj" and the ask the status with get status the display response is an empty "" project loaded.

i may be misinterpreting what you are saying, but it sounds like you are sending IP commands to BOTH servers? If so, that is incorrect usage if those two servers appear in the stage window of a single show. (A cluster is defined as all displays in the stage window of a single show.)

When operating in cluster mode all IP commands for load, transport (play, pause, goTo, etc  that typically require authenticate 1), input (timecode, MIDI, ArtNet), output (ArtNet, string-TCP/IP, etc), are only sent to the cluster master.

The cluster master is defined as the last display station to receive and successfully execute a load command. When a load command is requested, the cluster master looks in the show file for a list of all display stations in the stage window, and sets up communications with all of them. i.e. establishing cluster. Most communication between cluster master and its members are UDP.

The only time you communicate individually with a display server is for administrative functions, the ones typically requiring authenticate 2. (getMACAddr, powerDown, invoke VNC, etc)

Share this post


Link to post
Share on other sites

Hi Jfk and Thank you for your reply. No i do not spending double commands to different machines.I guess my horrible english and the hurry does not help, so i try  to be  more schematic: 

1. Why in your opinion at the start up, start only the server with   the Startup script and the second remain paused.

2. Why in your opinion if I send a tcp status of the machine that loaded the show it answer with “ “ ? Is it  Normal?

3. After the Windows re install i’ve got those problems. Is it a problem of eindows update? Do you think is a problem of a not correct tweking?

4. Do you think that a Fresh new installation and upgrade to Windows 10 can help?

 

thank you

 

 

 

Share this post


Link to post
Share on other sites

Now the last  updates from the  museum are that when the two servers starts, just the one with the start up script is playing, the other remain paused, than when the external device send the string to call the videos starts slightly out of sync.. i cannot understand how to fix it... help me please 😔

Share this post


Link to post
Share on other sites
13 hours ago, caleidocane said:

 

13 hours ago, caleidocane said:

 

I know English is not your strength, but there is a bit that is unclear ...

13 hours ago, caleidocane said:

Hi Jfk and Thank you for your reply. No i do not spending double commands to different machines.I guess my horrible english and the hurry does not help, so i try  to be  more schematic: 

1. Why in your opinion at the start up, start only the server with   the Startup script and the second remain paused.

2. Why in your opinion if I send a tcp status of the machine that loaded the show it answer with “ “ ? Is it  Normal?

Maybe, what is the exact string you are sending and the exact message you are receiving in response? A null is a valid response to some commands. Use command ID tagging to force a response.

Quote

3. After the Windows re install i’ve got those problems. Is it a problem of eindows update?

 I can not answer that question, that is the builders responsibility to solve through testing. Does not sound like it though.

Quote

Do you think is a problem of a not correct tweking?

Possibly. Could just as easily be related to hardware performance, codec choice / encoding settings,- movie resolution and frame rate, etc.

Quote

4. Do you think that a Fresh new installation and upgrade to Windows 10 can help?

Depends, did you build from a blank disc and than accurately follow the win 10 tweaking guide already -  if so, no.  Are you starting from machines that were purchased with Windows 10 on them, if so, yes. But I would not do that until other areas are eliminated.

Quote

 

12 hours ago, caleidocane said:

Now the last  updates from the  museum are that when the two servers starts, just the one with the start up script is playing, the other remain paused,

That is twice you have mentioned "remain paused". what does that mean?

Does it remain at the startup logo screen?

Does it load the show but does not run with the other server?

Quote

... than when the external device send the string to call the videos starts slightly out of sync.. i cannot understand how to fix it... help me please 😔

Starting out of synch may be related to how you are starting them. Please provide the exact command you are sending that results in " ... videos starts slightly out of sync ...". Are you triggering auxiliary timelines? and if so, are you managing their state changes? At may be as simple as adjusting how you start things.

 

Share this post


Link to post
Share on other sites
On 6/8/2019 at 10:29 PM, jfk said:

That is twice you have mentioned "remain paused". what does that mean?

 Does it remain at the startup logo screen?

 Does it load the show but does not run with the other server?

Yes on startup after the Watchout logos the server with the script go ahead the other remain paused at the first frame. If I change the location of the script, so if I delete the script from the current server ,and write the script on the other the problem is exchanged. 

 

On 6/8/2019 at 10:29 PM, jfk said:

Depends, did you build from a blank disc and than accurately follow the win 10 tweaking guide already -  if so, no.  Are you starting from machines that were purchased with Windows 10 on them, if so, yes. But I would not do that until other areas are eliminated.

it's windows 7, I've started from a blank iso with pen drive, and then followed every single step of the tweaking list for win7 founded in the forum. 

I've prepared numerous server and all still running without issues (with windows 10).

the strange thing that happens is that sometimes it runs perfectly, on Friday with one of the owners of the museum, for an extraordinary evening visit, it all started perfectly.

On 6/8/2019 at 10:29 PM, jfk said:

Maybe, what is the exact string you are sending and the exact message you are receiving in response? A null is a valid response to some commands. Use command ID tagging to force a response.

authenticate 1\r\ngetStatus\r\n

after the show load command it's start with the proper show loaded and then when the show starts (in the first server, while the other paused) if I send another getStatus the answer is ""and the other boolean value. sorry when the museum is closed I can transcribe all the response, I'll go the 17 June.

 

Thanks a lot

Share this post


Link to post
Share on other sites

Sure sounds like you have a network problem. Is there more than one NIC in your display servers?

Share this post


Link to post
Share on other sites
On 6/11/2019 at 11:13 AM, jfk said:

Sure sounds like you have a network problem. Is there more than one NIC in your display servers?

Thank you for your support! There is only one internet cable and two network ports, one of which is disabled from the control panel. 


I also tried to break away from the museum network and connect with a local switch, leaving the IP addresses set previously. I haven't tried to set a class c address (192 etc. etc), do you think it can make a difference?

Share this post


Link to post
Share on other sites
3 hours ago, caleidocane said:

Thank you for your support! There is only one internet cable and two network ports, one of which is disabled from the control panel. 


I also tried to break away from the museum network and connect with a local switch, leaving the IP addresses set previously. I haven't tried to set a class c address (192 etc. etc), do you think it can make a difference?

No, should not matter.

The fact that it runs correctly sometimes and acts up at others brings another possibility to mind. I have encountered network switches with automatic UDP storm / UDP flood / denial of service (DOS) protections. In those cases the switch interprets watchout's continuous UDP traffic as an attack and kicks in those protections (so it works for a while until the switch takes protective action). But that should not occur on a simple unmanaged local switch, although I did once encounter a D-Link unmanaged "secure" switch that included those types of protection.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...