View Issue Details

IDProjectCategoryView StatusLast Update
0016471MMASynchronizationpublic2021-09-20 13:36
Reporterrusty Assigned To 
PriorityhighSeveritymajorReproducibilityalways
Status feedbackResolutionreopened 
Product Version1.3.7 
Target Version2.1.0 
Summary0016471: Wi-Fi Sync: Playlists fail to update correctly when select by 'Random (refresh all)' is used
DescriptionIf Playlists are selected by 'Random Track (refresh all)' the tracks don't seem to sync correctly (or the Playlists update incorrectly). e.g. for the following playlists on the sync list:
Failed to sync (static; contains 17 tracks)
80's and 90's 3+ (contains 100 random tracks from the 80s/90's rated 3* or higher)
Good stuff 3+ (contains 100 random but non-classical/hebrew tracks that are rated 3* or higher)
Good stuff 4+ (contains 300 selected at random (refresh all) but non-classical/hebrew tracks that are rated 4* or higher)
Hebrew Good stuff (contains all hebrew tracks rated 3.5+; = 166 tracks)
Marvel DC (static; contains 36 tracks)
New Pop 2.5+ (contains 50 random but non-classical tracks from the year 2000+ rated 3.5* or higher)
Unrated - recent (contains 100 randomly selected (refresh all) unrated tracks, year 2000+)
Unrated - older (contains 100 random unrated tracks, year 1930-2000)

Wi-Fi Sync --> the following Playlist track numbers in MMA on the initial and 2 subsequent syncs):
Failed to Sync (17, 17, 17 tracks)
80's and 90's 3+ (99, 99, 99 tracks)
Good stuff 3+ (100, 100, 100 tracks)
Good stuff 4+ (297, 112, 127, tracks)
Hebrew Good stuff (166, 166, 166 tracks)
Marvel DC (36, 36, 36 tracks)
New pop 2.5+ (50, 50, 50 tracks)
Unrated - recent (100, 32, 30 tracks)
Unrated - older (40, 40, 40 tracks)

As you can see, for 2 of the playlists where tracks were selected randomly with 'refresh all', the number of tracks gradually gets further and further from the actual number based on what was selected in the sync list. Note that the, the total number of tracks after each sync operation was always over 900 (Total tracks 9xx, 927, 912), so it's hard to know whether new refreshed tracks are being synced but not to the playlists OR the new tracks aren't being synced and the old ones aren't being deleted, resulting in the lower playlist count.

I haven't tested select by 'Random Album (refresh all)'.
Additional InformationTest note: Peke please retest all of 0014553 after this is resolved.
TagsNo tags attached.
Attached Files
unrated-recent-playlist.jpg (13,797 bytes)   
unrated-recent-playlist.jpg (13,797 bytes)   
pixel2xl.mmdc (758 bytes)
nexus5x.mmdc (738 bytes)
Fixed in build2239

Relationships

related to 0014553 closedLudek MMW 5 Make auto-playlists with random order more persistent 
related to 0014761 closedLudek MMW 5 Cloud tracks (not downloaded) cannot be Wi-Fi synced to MMA 
related to 0012509 closedLudek MMW v4 Sync of Playlist with stream URLs has sync fail 
related to 0013290 assignedmartin MMW 5 Online YouTube tracks fails to sync to device 
related to 0012721 closedmarek MMA Failed conversion error appears when it shouldn't  
related to 0014374 closedpeke MMW v4 WiFi-Sync: Incomplete/slow sync when some network files are inaccessible 

Activities

Ludek

2020-04-03 10:46

developer   ~0057472

Last edited: 2020-04-03 10:56

I am unable to replicate the described issue.

Some questions:
1) I supose that you have "Delete unselected content copies" enabled and used between each sync, right?
2) Were there any errors re inaccessibility of certain files? e.g. dead-links ?
3) What is your exact sync profile settings? Could you export it as MMDC ?
4) Which MMA version was used in the test ?
5) Does this happen also when only "Unrated - recent" auto-playlist is synced or the others (auto-)playlists needs to be synced too for the bug to appear?
6) Could you attach debug log? Ideally with as smaller set of files and playlists as possible -- debug logs with hundreds of files are hard to analyze.

rusty

2020-04-03 23:17

administrator   ~0057478

To answer your questions:
1) Yes--track deletion of non-library files was enabled
2) Yes--there was 1 file '50 ways to leave your lover' that repeatedly failed to sync. I have 2 versions of it, one is stored to a Google drive--presumably that was the one that was failing to sync.
3) MMDC files attached (Pixel2xl was used for the initial set of tests yesterday, Nexus5x for the tests performed today (see below).
4) MMA 1.3.8.912 was used.
5) I'm not sure. Today when I retested, the issue only seemed to occur when I synced overlapping playlists, however, I was never able to get it to occur as frequently (nor in a pattern of increasing amounts of failures) like yesterday.
6) Attached

The tests below were performed by creating a smaller subset of the playlists synced yesterday, each containing 15 tracks:
Good stuff2 3.5+ (has tracks that overlap with Good Stuff2 4+)
Good stuff2 4+ (refresh all)
Unrated - recent2 (refresh all)
Unrated - older2

The bug only occurred on the 5th sync operation.

Steps:
Sync1 of unrated-recent and unrated-older 0-28669
Log ID: ZTE81SRNH1
-->success
Sync2 28671-45651
Log ID: UYUA16WR5L
--> success (15 tracks deleted, and 15 downloaded)
Sync3 46000-60726
Log ID: MD132HWX1I
-->Success (14 tracks deleted, and 14 downloaded)
Added Overlapping Playlists to the sync list (GoodStuff2 3.5+ and Goodstuff2 4+: each limited to 15 tracks but there are likely to be tracks that overlap)
Initiated Sync4: 61783-89653
Log ID: 4UXYBWIC3Q
--> Success (14 tracks deleted, and 42 downloaded)
Initiated Sync 5: 889654-111760
Log ID: UZDTEP4THS
--> ?BUG? only 14 tracks in unrated-recent2 (26 deleted, 27 downloaded)
Initiated Sync6: 111761-136567
Log ID: H6CBZG498E
--> Success (28 deleted, 29 downloaded)
Edited track stored to Google Drive to Genre=Acoustic Rock
Initiated Sync7: 137006-160017
Log ID: D88I293PI9
--> Success (30 deleted, 30? downloaded)
Initiated Sync8: 160018-182295
--> Success (28 deleted, 28 downloaded)
16471.7z (1,582,692 bytes)

Ludek

2020-04-04 18:33

developer   ~0057485

Last edited: 2020-04-06 16:39

I analyzed the logs and I can confirm that the track "50 Ways to Leave Your lover" that was missing came into the play on 5th sync.
The "50 Ways to Leave Your lover" is could track from GoogleDrive (MM db id: 27635, Google Drive id: 1zC4NxijUmfNz5bCeCtILszCjx_dYQHa2 )

So the key factor to replicate seems to add some cloud only tracks on the sync list.
EDIT1: After re-analyzing: Although the cloud track "50 Ways to Leave Your lover" was in the log around the 5th sync -- it was not on the "sync-list" that time, so there might be another problem, to be localized...

To be found:
1) why the cloud track was not paired with its local copy
2) why (even if not paired) was not copied from Google Drive to MMA -- there is already a mechanism for this implemented
EDIT2: found that the mechanism of transfering track from cloud location to MMA was implemented just for USB sync, to be implemented for Wi-Fi sync too. Tracked as 0014761

Ludek

2020-04-06 16:41

developer   ~0057496

Last edited: 2020-04-06 17:45

Assigned to Martin to analyze UZDTEP4THS whether he can see something , in MMW log I see that unrated-recent2 had 15 files -- so I don't know why it would show only 14 in MMA

EDIT:
Assigned back to me, the playlist really had only 14 items -- line 00111580 in the MMW log.

It is most probably caused by 0012509 where a track with URL path got to the playlist and was filtered out (due to 0012509)
Another possibility is that it was YouTube track ( 0013290 ), but also could be temporarily inaccessible network track ( 0014374 ) or the cloud track ( 0014761 )

Added more dubug info for the next MM5 build so I will see the exact reason next time in the MM5 log -- anyhow I guess that MMA must reported this as a track download failure?

rusty

2020-04-07 21:11

administrator   ~0057525

Moved to MMA since the only problem I've been able to replicate thus far is that MMA can, in some cases, show tracks as having successfully synced even when they haven't.

Ludek

2020-04-17 09:23

developer   ~0057611

Rusty, please generate MMW log (with accompanying MMA log) so that I can see what kind of tracks fails to sync and why (I have added more debug info in 2239+)