Plugin: SJQv4

[Slugger]

Right... wow it's been so long since I've touched any of my SJQ scripts that I've forgotten how I even implemented this stuff. I think it's more to do with how I wish I had done it and how I would have done it for SJQv5 had that ever seen the light of day.

Anyway, yeah, the API variables are actually injected into the scripts automatically so you don't need to import anything.

But you don't prefix with "Global". See here. I actually had to go skim over this to remind myself.

So to iterate through the system messages and delete specific ones, I might do something like this (all completely untested):

// Find and delete every system message that contains the word 'Halt' in its message text
SystemMessageAPI.GetSystemMessages().findAll { SystemMessageAPI.GetSystemMessageString(it).contains('Halt') }.each {
   SystemMessageAPI.DeleteSystemMessage(it)
}

[Fuzzy]

Actually, i seemed to have sorted it out as you were typing that it seems. Here's what I ended up with:

if(SJQ4_METADATA.get("SJQ4_MSG_TYPE")=="Halt Detected in Recording") {
	if(SJQ4_METADATA.get("SJQ4_TEXT").contains("VIP-211-1")) {
		if (Integer.parseInt(SJQ4_METADATA.get("SJQ4_COUNT")) == 2) {
			for ( obj in SystemMessageAPI.GetSystemMessages() ) { 
				if(SystemMessageAPI.GetSystemMessageString(obj)==SJQ4_METADATA.get("SJQ4_TEXT")) {
					SystemMessageAPI.DeleteSystemMessage(obj); 
				}
			}
			SystemMessageAPI.PostSystemMessage(1204,2,"R5000HD Software was reset due to : " + SJQ4_METADATA.get("SJQ4_TEXT"), null);
			return 0;
		} else {
			return 2;
		}			
	} else {
		return 2;
	}
} else {
	return 2;
}

only current issue is that, while yes, this does delete any previous messages that led up to the reset, and it does generate a warning level message to say that it reset it, the reset itself generates a new halt message afterward, that doesn't get removed. I'll have to add on yet another section to this pretest script i think to handle that - if I feel the need.

[bikesquid]

I must have screwed up something unknowingly. I've got a few tasks that simply are stuck. Not active, the associated .exe's aren't running, so I know their are dead, but sjq won't let me delete them and won't go on to do other work. I've tried a restart of sage as well with no luck. Any suggestions on this greatly appreciated, I'm lost.

[Slugger]

You might have to remove the entries from the db manually:

https://code.google.com/p/sagetv-add...ModifyDatabase

[bikesquid]

I did try that, just did again hoping for a different outcome. What I get is if I do a query in the console of "SELECT * FROM QUEUE " I get get the full queue list, as expected, so far so good. If I run the various commands from your post #775, such as "DELETE * FROM queue WHERE state = 'WAITING'" I get "DELETE * FROM queue WHERE state = 'WAITING';
Syntax error in SQL statement "DELETE *[*] FROM QUEUE WHERE STATE = 'WAITING' "; expected "identifier"; SQL statement:
DELETE * FROM queue WHERE state = 'WAITING' [42001-145] 42001/42001 (Help)" as the output.

When I do "DELETE * FROM queue" I get the same sort of output.

What I'd really like to do is just clear the queue and remove all tasks and clear the clients assigned tasks. Then I can add back what I want.

[Slugger]

Ok, I updated 775, the SQL syntax is wrong.

Drop the * in the queries, then they should work.

So DELETE FROM queue WHERE state = 'WAITING'

[bikesquid]

Originally Posted by Slugger

Ok, I updated 775, the SQL syntax is wrong.

Drop the * in the queries, then they should work.

So DELETE FROM queue WHERE state = 'WAITING'

It's different... but not working yet:
DELETE FROM queue WHERE state = 'WAITING';
Row not found when trying to delete from index "PUBLIC.FK_CLIENT__QUEUE_INDEX_1: ( /* key:33352 */ 33352, 'CUT', TIMESTAMP '2013-11-04 16:30:27.717', NULL, NULL, 'WAITING', NULL, NULL, FALSE)"; SQL statement:
DELETE FROM queue WHERE state = 'WAITING' [90112-145] 90112/90112 (Help)

[Slugger]

Your database file is corrupted. If you have a backup, restore it. If not, I'm afraid you're going to have to delete the entire database and start fresh.

[bikesquid]

Originally Posted by Slugger

Your database file is corrupted. If you have a backup, restore it. If not, I'm afraid you're going to have to delete the entire database and start fresh.

No problem, just the sjq4.h2.db file?

[Slugger]

You must stop SageTV first then go to the following dir and remove the following files and subdir (the .lock.db file probably won't be there, that's ok).

 Directory of C:\Program Files (x86)\SageTV\SageTV\plugins\sjq

11/05/2013  11:05 AM         2,889,728 sjq4.h2.db
11/05/2013  04:12 AM    <DIR>          sjq4.lobs.db
11/05/2013  03:31 AM               103 sjq4.lock.db

[bikesquid]

Thanks again Slugger, that did it! Wonder what happened...glad my backups are working at least!

[TechBill]

I had to reiinstall SAgeTV and I cannot remember how did I set up to use Groovy script to restart SageTV when there no recording going on.

I remember using stand alone Groovy exe and I download it but it giving me a license not found error when I tried to set it up.

Is there a new way of restarting SageTV nightly using this script or still need to use stand alone exe?

Still struggling trying to make this work again

Can anyone here help me out ?

Thank you

Bill

[Slugger]

Follow this tutorial, except use the script below instead of the mentioned exe script (you can use the mentioned test script as is).

Utility.ServerRestart()
sleep 300000
return 0

Just make sure the sleep time is long enough for Sage to completely restart. That value is currently set at 5 minutes (300 seconds/300000 ms). It still must be run on an external task client. You don't need SageGroovy other than to test scripts, etc. If you want to use SageGroovy copy the salicense.jar file from SageTV\JARs\ into the lib dir on SageGroovy, that will fix the licensing errors from it. The standalone SJQ task client does not use licensing.

[TechBill]

Originally Posted by Slugger

Follow this tutorial, except use the script below instead of the mentioned exe script (you can use the mentioned test script as is).

Utility.ServerRestart()
sleep 300000
return 0

Just make sure the sleep time is long enough for Sage to completely restart. That value is currently set at 5 minutes (300 seconds/300000 ms). It still must be run on an external task client. You don't need SageGroovy other than to test scripts, etc. If you want to use SageGroovy copy the salicense.jar file from SageTV\JARs\ into the lib dir on SageGroovy, that will fix the licensing errors from it. The standalone SJQ task client does not use licensing.

Alright I will give this method a try

Thank you

Bill

[bikesquid]

I started a new thread, but maybe it should be here instead as I'm thinking it's related to sjq pretty directly....http://forums.sagetv.com/forums/showthread.php?t=61004

[Slugger]

The connections from SJQ are normal. They will timeout and clean themselves up based on the TCP settings on your system. This is usually like 5 minutes or so. But that seems like a lot of active connections, but that could be normal based on the number of jobs you have setup in SJQ. If you're running Windows 7 then this may or may not help:

https://www.google.ca/url?sa=t&rct=j...57155469,d.aWM

Note: Just a quick google search, I have no idea if that's an issue or is related, etc. Use at your own risk.

[Slugger]

"Normal" meaning that Windows should clean things up in a timely manner. It's probably a bug in the SJQ task client or server code where one side is closing the connection and the other side isn't explicitly acknowledging the close, which is probably why the socket remains open until Windows times it out and cleans it up.

At this stage of the game, a fix is very unlikely (the current code has been out there since June 2011 so it hasn't been a widespread problem/my familiarity with the code at this point is low, to say the least), but if you provide the stack trace of the connection refused errors you're seeing I will try to diagnose the issue and if it's a critical bug then I might release a fix.

[bikesquid]

That's helpful, I was starting to go that route, but saw this large number of open ports for sagetv. There are a running (approx) 30 open ports waiting, yes they time out but new ones pop up at the same rate. current task queue shows 2 jobs running, 6 waiting. The completed tasks is closer to 30 though....

I've also noticed that the HD300's keep opening ports related to something else running on the same box, so they also leave lots of open/waiting ports. It's probably a perfect storm of danglers that clears up, but only after the damage is done.
Thx.

[bikesquid]

Originally Posted by Slugger

At this stage of the game, a fix is very unlikely (the current code has been out there since June 2011 so it hasn't been a widespread problem/my familiarity with the code at this point is low, to say the least), but if you provide the stack trace of the connection refused errors you're seeing I will try to diagnose the issue and if it's a critical bug then I might release a fix.

Has it been that long?? 2011

Understood that it's EOL and probably not going to see a fix. But it's possible I'm doing something in my scripts that's exploiting a problem that I can do differently?

If I knew how to get any more info related to stack trace I'd be happy to provide, the only think I find via a web search is "Thread .currentThread().getStackTrace()." which I suppose I could throw in a suspect script with a Println which would get it in the task log file, but I'm not sure that's what you mean. Maybe you're suggesting a netsh trace?

[Slugger]

You said you were getting connection refused errors. I assume where ever you saw that error would also have a stack trace attached to it. Where are you seeing those errors logged?

And the last SJQ server release was indeed on June 20, 2011; the last task client release was Oct 18, 2011. Boy how time flies.