Remote JACE Reboot

[tridiumtech]

A potential client wants to monitor/control a very remote facility. They like Tridium & they want to use
a JACE to monitor the facility energy status + some other control points. They know that the
JACE has a 'Engine Watchdog Policy' that has the ability to detect and deal with a 'hung' or 'stalled'
station (perform a auto reboot and auto start).

But, once in awhile, sometimes 'strange' happens and you can't force the JACE reboot remotely (using the
Application director) or the remote reboot doesn't solve the problem. Or the VM got 'killed'. Or
planet Jupiter is square with Uranus - anyway, something 'strange' happens and the only way out is to
manually power reboot the T-Box. This is what the client is concern about - it just not easy to drive out
to the remote site and do a power or 'hard' reboot. They want a way to reboot with out going out to the site
(and they'll pay for it - $$$; got to love those military guys).

We talked about putting in, don't laugh, an analog phone line so you could dial in to a remote modem
which would control a 'black box' that would power reboot the JACE. Another idea was to put in a backup JACE,
but then there's the issue of what happens when #2 JACE gets 'hung'?

Another idea: write a utility (System Monitor Service?) that monitors heaps & memory usage to track the trend
of memory leaks (generate alarms) & to force a reboot based on a schedule (from Niagara-Central).
But if the VM is dead, it's dead - I'm not the complete Niagara whiz, but no super utility code is going to save
the day if the VM is killed. Right?

Anybody got any ideas?

[MonitoringGuy]

Do a Google search for "WebRelay". There are several different manufacturers of similar products, but that would give you an idea of what is out there....

[klrogers]

I have had a Jace that was so low on heap that I could not maintain a connection with WB or force a reboot from the platform application director, but was able to reboot the Jace using the very unsupported web commands. If you need more infomation on them PM me, not something that should be placed on a public forum.

Kevin

[tridiumtech]

I saw a post on Niagara Central that went into the realm of 'unsupported web commands' - left that idea out -
trying it keep this thread "GP" rated ;-) But thanks. Good comments so far.

[control$]

A potential client wants to monitor/control a very remote facility. They like Tridium & they want to use
a JACE to monitor the facility energy status + some other control points. They know that the
JACE has a 'Engine Watchdog Policy' that has the ability to detect and deal with a 'hung' or 'stalled'
station (perform a auto reboot and auto start).

But, once in awhile, sometimes 'strange' happens and you can't force the JACE reboot remotely (using the
Application director) or the remote reboot doesn't solve the problem. Or the VM got 'killed'. Or
planet Jupiter is square with Uranus - anyway, something 'strange' happens and the only way out is to
manually power reboot the T-Box. This is what the client is concern about - it just not easy to drive out
to the remote site and do a power or 'hard' reboot. They want a way to reboot with out going out to the site
(and they'll pay for it - $$$; got to love those military guys).

We talked about putting in, don't laugh, an analog phone line so you could dial in to a remote modem
which would control a 'black box' that would power reboot the JACE. Another idea was to put in a backup JACE,
but then there's the issue of what happens when #2 JACE gets 'hung'?

Another idea: write a utility (System Monitor Service?) that monitors heaps & memory usage to track the trend
of memory leaks (generate alarms) & to force a reboot based on a schedule (from Niagara-Central).
But if the VM is dead, it's dead - I'm not the complete Niagara whiz, but no super utility code is going to save
the day if the VM is killed. Right?

Anybody got any ideas?

When doing critical system requiring redundant controls, I use the "heatbeat" system. Using another control to simply poll that device at fix time.(usually a hard wired point but network polling works). If this device does not receive a status after x amount of tries, have it open a relay to kill the power to the jace. Thus forcing a reset.

[xarralu]

When doing critical system requiring redundant controls, I use the "heatbeat" system. Using another control to simply poll that device at fix time.(usually a hard wired point but network polling works). If this device does not receive a status after x amount of tries, have it open a relay to kill the power to the jace. Thus forcing a reset.

Yea that is what we do. A delay on make and a delay on break. Have it go out on a network point to another module and if it fails on one or the other, it opens the relay for XX seconds and then waits XX minutes before it tries it again.

[Shermer321]

I am running into this right now, can someone PM me what to do to possibly reboot via the web?

I have had a Jace that was so low on heap that I could not maintain a connection with WB or force a reboot from the platform application director, but was able to reboot the Jace using the very unsupported web commands. If you need more infomation on them PM me, not something that should be placed on a public forum.

Kevin

[MaxBurn]

Can I run something on the station to reboot it? This is not a lock up or unresponsive situation, I just have a need to reboot the station.

[klrogers]

Can I run something on the station to reboot it? This is not a lock up or unresponsive situation, I just have a need to reboot the station.

Right Click on Platform Services/Actions

[MaxBurn]

Right Click on Platform Services/Actions

I meant more along the lines can I make some logic to do that, that is triggered by something else.

[klrogers]

I meant more along the lines can I make some logic to do that, that is triggered by something else.

The SysInfo module in the Ax Community module as a reboot program, that should show you how to write something yourself.

[MaxBurn]

The SysInfo module in the Ax Community module as a reboot program, that should show you how to write something yourself.

I don't suppose there is a N4 equivalent?

[klrogers]

I don't suppose there is a N4 equivalent?

There is check out: https://sourceforge.net/projects/niagaraaxcommun/files/

To entry is N4 version

[MaxBurn]

There is check out: https://sourceforge.net/projects/niagaraaxcommun/files/

To entry is N4 version

Thanks, that worked.

In fact it worked a little too well, I threw a boolean in and turned it on and it rebooted. Then it rebooted again. Nice.

[mwitt]

Thanks, that worked.

In fact it worked a little too well, I threw a boolean in and turned it on and it rebooted. Then it rebooted again. Nice.

OneShot would help here right?

[MaxBurn]

OneShot would help here right?

That's what I ended up doing. I was concerned with not knowing exactly how fast the restart triggers and the oneshot not having time to complete. I guess "fortunately" N4 on a jace 6 is slow. So in practice the oneshot set to a half second does have enough time to complete. But I also had to loop the output around to the boolean input so that it resets the boolean I used to trigger the oneshot. Otherwise it would just trigger the oneshot again on the next boot doing the same boot loop.

On the other hand I got a little practice with restoring backups.

[szeszycki12]

do you still have record of those unsupported reboot commands?

[Elbanhawy]

if the Jace lost WIFI and connection to the station for example, and I need to reboot it from within the JACE, how do I do that?