Plugin: SJQv4

[phelme]

Originally Posted by Slugger

Not sure why the RMI service would suddenly stop working. You might want to track down stuckless and see if he has any thoughts (sagex-services is his plugin). My initial guess is that something else is latching onto port 1098? Only a guess though.

After some more snooping, you're right, it's another process. The PlayOn MediaMallServer is also using 1098. Not sure why that suddenly started happening, been running that for ages. Unless a recent update from them made some port changes. Thanks!

[MrD]

I am seeing the server PING a remote client every few hundred milliseconds even though the server is set to ping the client every 600 seconds.

[Listener] Received connection
[Handler] CMD PING 
[Handler] CMD QUIT

over and over again in the remote client (Windows XP sp2) log, server is linux.

is there any reason this is occurring?

[Slugger]

Originally Posted by MrD

I am seeing the server PING a remote client every few hundred milliseconds even though the server is set to ping the client every 600 seconds.

[Listener] Received connection
[Handler] CMD PING 
[Handler] CMD QUIT

over and over again in the remote client (Windows XP sp2) log, server is linux.

is there any reason this is occurring?

Can't think of a reason, other than a bug. Is there a task in the queue stuck in waiting state? Disable the server plugin, does it stop? If so, restart the server plugin, does it start again at that interval? If so, there's a bug somewhere.

[MrD]

I deleted a handful of queued tasks that may have been causing problems b/c they were test groovy scripts and things seemed to have calmed down.

I am interested in understanding how that would be related to the PING flood I was seeing?

[Slugger]

Most likely a bug. I'd need to see the server logs corresponding with those client logs.

Alternatively, if you had tasks stuck in the queue with the test script constantly returning 1 (return to queue and retry later), well depending on how many tasks were stuck in that situation, you might end up with this because a task is returned to the queue and then that triggers a queue reprocess, which would cause the other tasks to retry (just not the one that was just sent back). With that said, the behaviour does sound a little buggy, but if cleaning up the queue fixed the situation then it's not an issue I'm likely to investigate too deeply at this point in time.

[MrD]

Cool beans, makes sense to me.

Read your other post... being a software engineer myself, I know exactly where you are.

Well in 4 months it will be 12 mos since the buyout, I'm positive something will change then good, bad or ugly.

[firenice]

My server has been running great until Saturday. For some reason a task or task got control and filled my WHS system drive with lobs.db files.

That said, I cleaned the .lobs.db folder out and disabled the task client on the server. This has eliminated the lobs.db from being created but....

The server task client in the plugins setup now reports "fail". This is how I knew to disable it. I thought maybe an uninstall and reinstall would straighten it out but no.

I can open the SJQ4 in the UI but no clients show up. It has not run a scheduled task in a few of days and I am thinking maybe a core SJQ4 file may have been corrupted.

I have logs but they are very large and I can cut em down if you can tell me any paticular items to look for.

Thanks in advance.

Mike

Thanks.

[Slugger]

Originally Posted by firenice

My server has been running great until Saturday. For some reason a task or task got control and filled my WHS system drive with lobs.db files.

That said, I cleaned the .lobs.db folder out and disabled the task client on the server. This has eliminated the lobs.db from being created but....

The server task client in the plugins setup now reports "fail". This is how I knew to disable it. I thought maybe an uninstall and reinstall would straighten it out but no.

I can open the SJQ4 in the UI but no clients show up. It has not run a scheduled task in a few of days and I am thinking maybe a core SJQ4 file may have been corrupted.

I have logs but they are very large and I can cut em down if you can tell me any paticular items to look for.

Thanks in advance.

Mike

Thanks.

The usual cause of this problem has been tasks stuck in the queue with buggy test scripts associated with them.

First, clean up your task queue as described here:

http://code.google.com/p/sagetv-addo...ModifyDatabase

You'll want to look at the QUEUE table once connected and remove any tasks in WAITING or RETURNED state. If there are plenty in a RETURNED state then you'll want to figure out why (i.e. usually a buggy test script). It's these tasks that constantly requeue and reexecute that end up usually being the cause of .lobs.db filling up.

[firenice]

Originally Posted by Slugger

The usual cause of this problem has been tasks stuck in the queue with buggy test scripts associated with them.

First, clean up your task queue as described here:

http://code.google.com/p/sagetv-addo...ModifyDatabase

You'll want to look at the QUEUE table once connected and remove any tasks in WAITING or RETURNED state. If there are plenty in a RETURNED state then you'll want to figure out why (i.e. usually a buggy test script). It's these tasks that constantly requeue and reexecute that end up usually being the cause of .lobs.db filling up.

OK Thanks. Cleaned the returned and waiting out but now this is still happening.

I go into the SJQ UI and no clients are there. Capture attached. I also attached the last few lines of the SJQ log. Seems to be a communication error but everything seems to be working except this and my tasks.

Thanks again
Mike

[Slugger]

Those (specific) IOErrors are normal (at least expected). Register the task client again. Check the db clients table and see if any are listed, but it looks like you've managed to unregister all task clients. Registering them again should bring them back online.

[firenice]

Thanks Slugger.
Attempted to setup clients again but it failed. The server client, when I attempted to enable it reported the plugin was corrupt and to uninstall and reinstall. I did prior to last post and it still did not show nor allow any clients to be added.
Thanks for you help
Mike

[Slugger]

Any errors (exceptions) in the sage debug log?

[firenice]

Originally Posted by Slugger

Any errors (exceptions) in the sage debug log?

No, no exceptions. My client task client was in the database but the server task client wasn't. Strange because the client isn't showing up.

I think this may be related to a commutation error with my server rather than a SJQ error. I have had some other issues and now playon is not showing my Netflix que either. Think maybe a port got closed or something that all this stuff relies on.

Appreciate your help and I'll get back with you if after I work through all of this.

Thanks
Mike

[firenice]

Slugger

Just a follow up to my no client problem.

I did have another issue. I discovered that a user profile was corrupt on my server and that was the main issue.

I did have to reregister one client but it picked it up right away afterwards.

Thanks again for the assist

Mike

[Slugger]

Effective immediately, official support for this plugin has ended. More details available here.

A special thank you to all who have donated to this project. Because of your support, a donation for a little over $500 has been made to the Canadian Cancer Society. Again, thank you for your support!

[MattHelm]

Help, not a programmer at all (well a little Quick BASIC & AREXX). Never could get a handle on Object-oriented programming. I'm OK with Procedural programming and Object-oriented data, if I have code for examples, but OOP fries what little brains I have left, and even zombies don't like fried brains!

Could someone write me a few SJQv4 scripts?

1. Once a day, starting when no recordings are scheduled for 2 (or how ever many) hours, start a windows scheduled task. (at least I think you can make a task that won't run on it's own, but will run when Schtasks.exe calls it)

2. Run VideoRedo to do a QSF and repackage from a .TS to a .MKV. (I still have to move my VideoRedo license from my "main" machine to my server)

Anyone????? Please!!!!

(I hope if I have 2 examples like these, I can hack any future out by myself. I'm OK modifying code someone else wrote.)

[graywolf]

Can help with part of #1.

Below code is the test script for the backup job. Checks if a Scheduled Recording will occur in reqdMins (set to 45 in my script). If it is on LiveTV (use it as background/white noise for sleeping in MBR) then go ahead an perform the task. It only requeues if something is recording or will record in the specified time.

If you ever get the other tasks worked out, I would probably be interested in checking them out.

/*
    SageTV backup tool tester (i.e. is it "safe" to stop and backup SageTV?)
    Last Modification: 2011-11-21
               Original Author: Derek Battams <derek AT battams DOT ca>
                              
    Use this script as the test for the full_backup.groovy script.  Basically this
    script just checks a bunch of conditions to determine if it is "safe" to shutdown
    SageTV and perform the app dir backup.  Remove/add/modify conditions to your needs
    below.
    
    Remember, return 1 to push the task back to the queue, 0 to tell SJQ to run the backup script
    or 2 to tell SJQ to skip the backup.
*/

def now = System.currentTimeMillis()
def reqdMins = 45L
int ScheduledRecording = 0 ;

println " ";
Global.GetScheduledRecordingsForTime(now, now + (reqdMins * 60000L)).each { mfAiring ->
    if( AiringAPI.IsNotManualOrFavorite(mfAiring) ) { 
       println (ShowAPI.GetShowTitle(mfAiring) + " is LiveTV - no issue");
    } else {
       ScheduledRecording = ScheduledRecording + 1 ;
       println ( ShowAPI.GetShowTitle(mfAiring) + " is Scheduled to record within $reqdMins minutes") ;
    } 
}

if ( ScheduledRecording > 0 ) {
    // Scheduled Recording upcoming, Return for when it is safe
    println ( "Returned to Queue until safe to process");
    sleep 900000
    return 1
} else {
    // Looks to be safe, let's do this!!
    println ("Safe to kick off Backup job. No recordings scheduled within $reqdMins minutes");
    return 0
}

[MattHelm]

Originally Posted by graywolf

...
Below code is the test script for the backup job. Checks if a Scheduled Recording will occur in reqdMins (set to 45 in my script). If it is on LiveTV (use it as background/white noise for sleeping in MBR) then go ahead an perform the task. It only requeues if something is recording or will record in the specified time.

If you ever get the other tasks worked out, I would probably be interested in checking them out.
...

Thanks. Been busy since you posted, so haven't tried it yet, but thanks!!! I will post ANY scripts I come up with!!!!! Sharing is good!

BTW, will this run if someone is watching SageTV?

[wayner]

@MattHelm - I have code that calls comskip and Handbrake from SJQV4. While this is not exactly what you are looking for it may be helpful since it shows you how to call an external process and handle the output, which can be tricky with verbose software and Handbrake and Comskip are both verbose in their output.

I use VideoReDo but I have never used it in batch mode with automated processes. However it may make sense for me to use it in place of Handbrake to convert files to an iPad friendly format (like .mp4 or .m4v) and cut out commercials (I process all kids' files which are originally MPEG2 - they get comskipped, converted to .mp4 and moved to a show-specific folder). Have you done any stuff with VRD in batch mode before?

[graywolf]

Think you would need to put in another check to see if someone was watching something. Mine only checked Currently/Soon to start recordings.

if( AiringAPI.IsNotManualOrFavorite(mfAiring) ) means it is LiveTV currently showing/"recording".