EPG123 guide display changes cause Long Display Time?

An evolving, supported alternative to Rovi
Forum rules
★ Download the latest EPG123 here: https://garyan2.github.io/ <> Setup guide here: https://garyan2.github.io/install.html
User avatar
garyan2

Posts: 7480
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#21

Post by garyan2 » Thu Apr 30, 2020 11:53 pm

TGBFAN wrote: Thu Apr 30, 2020 9:35 pmGary, my mcupdate_scheduled task (which has never been altered or edited) does not contain "-dbgc" or "-uf"

How is the every-five-day garbage collection accomplished without these switches included in the mcupdate_scheduled task?
If you look at the table I posted above, you will see that the default configuration of mcupdate is to do the maintenance task (fNoMaintenance is default false). mcupdate will check to see if maintenance is due, and if so, kick off another mcupdate process to presumably run "-uf -dbgc", or something equivalent. I haven't dove that deep into the mcupdate code.

So any execution of mcupdate will potentially start a client maintenance session (unless it has the -updateTrigger and/or -noMaintenance switches) and the mcupdate_scheduled task is scheduled to run daily. If you disable that task, then there is nothing else calling mcupdate that would check and/or perform the maintenance of your database.

EDIT: nothing else except the PVR reindexing task EPG123 does after it imports the mxf file. I'll have to run an experiment to see if the indexing task could kick off the maintenance.
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

stuartm

Posts: 723
Joined: Mon Nov 05, 2012 8:05 pm
Location: Longmont, CO

HTPC Specs: Show details

#22

Post by stuartm » Fri May 01, 2020 12:24 am

I doubt the PVR Reindexing task kicks of the garbage collect since on my machine with the mcupdate_scheduled disabled that would still get run by EPG123 and the DB was not being garbage collected.

Space

Posts: 2841
Joined: Sun Jun 02, 2013 9:44 pm
Location:

HTPC Specs: Show details

#23

Post by Space » Fri May 01, 2020 12:35 am

redmed wrote: Thu Apr 30, 2020 6:10 pm It's been four days since I ran the garbage collection and the guide display is still fast! Thanks Gary!

Something I have noticed the last few months is that after a guide load i will do a "view scheduled" and see what the last day scheduled is. Then open the guide for the first time and the guide will not display the last day that shows with "view scheduled". It will be a day or two short. I will then close the guide and then open the guide for the second time then the guide will show the same last day that showed in "view schedule". There was a difference I've never seen before after this tuesday morning load. The "view scheduled" showed the schedule until 5\11 but the guide would only show up to 5\10 no matter how many times I open & closed the guide. This had me worried that the garbage collection did not go as smoothly as I had thought. But this morning after the thursday load the "view schedule" showed the schedule until 5\13. The first open of the guide only went to 5\12 but the second time I opened the guide it opened to 5\13 matching the schedule again. Does opening the guide trigger additional guide processing?
This is a bug with WMC that has always happened for me (even with Rovi/Microsoft data, and even before that when it used Gracenote). It is nothing new (at least for me).

When this happens, the guide just "stops" at a certain date/time and you can no longer scroll to the right. You will not see the "No data available" for the further out to the right timeslots that you would normally see. This is indicative of it just being a GUI/display problem, not a problem with the underlying data, which is still there, just not displayed by the GUI.

I've found two ways to "clear" the problem;

1) Run another guide update (with the old MS guide data it would return immediately since there would be no new data to download). I'm not sure if the same would happen if you did another EPG123 update, but I assume it would because, as you saw, it was fixed the next day you looked at it (most likely because of the second update). It is probably related to one of the two processes that are run after the update completes, and not the update itself.

2) Exit the WMC GUI and restart it.

Note that occasionally exiting WMC and restarting would not fix the problem, but I never had it not be fixed by starting a manual update of the guide data.

Now that we no longer use the built-in update mechanism (and instead use the external EPG123), I wonder if telling WMC to redownload the guide data would still work to fix this problem. If you try to update the guide using this old mechanism, it appears to do nothing, but it may still run the processes that tell it to update the guide "end" so that it scrolls further out to show the whole guide.

User avatar
garyan2

Posts: 7480
Joined: Fri Nov 27, 2015 7:23 pm
Location:

HTPC Specs: Show details

#24

Post by garyan2 » Fri May 01, 2020 2:50 am

stuartm wrote: Fri May 01, 2020 12:24 am I doubt the PVR Reindexing task kicks of the garbage collect since on my machine with the mcupdate_scheduled disabled that would still get run by EPG123 and the DB was not being garbage collected.
Yah, I think you're right. I just ran EPG123 on a VM with mcupdate_scheduled disabled and it didn't do anything. Maybe it is the -pscn switch that does it. If I have time, I'll delve into the mcupdate code again, but not a high priority.
- Gary
Keeping WMC alive beyond January 2020. https://garyan2.github.io

WMCLives

Posts: 54
Joined: Fri Jan 03, 2020 3:54 am
Location:

HTPC Specs: Show details

#25

Post by WMCLives » Fri May 01, 2020 1:33 pm

The EPG123 install info was great. Up and running very quickly. I notice a lot of discussion about which tasks should be running. I try to keep my machine clean and running without extra tasks. Maybe when the answers are found the install info could also say which MCE tasks need to run and which can be disabled.

Post Reply