Sqlite3: Sleeping for 200ms to retry busy DB

[loudbutt]

Server Version#: v.1.41.8.9834-071366d65
Player Version#: N/A

I’m getting the dreaded “Sqlite3: Sleeping for 200ms to retry busy DB.” message over and over and over again to the point where PMS isn’t even usable.

I’m running PMS as a docker container within Unraid. Docker and all the non-media files are run off a NVME SSD, so it’s definitely not a hard drive speed issue. I tried repairing it with the tool by ChuckPA and it didn’t accomplish anything.

I know it’s not a server issue or that the db is excessively large. I spun up an Emby instance last night, it was able to index the entire library overnight and everything works perfectly when I tried it this morning.

I tried restarting PMS many times yesterday to no avail. I’m just seeing the Sqlite3 error message over and over and over again.

I’ve attached my server logs that are set to debug

Plex Media Server Logs Jun 24 2025.zip (3.7 MB)

[ChuckPa]

While the AMD Epic is a fast CPU (many threads), the per-thread CPU speed is nominal.

AMD EPYC 7D12 Benchmark

I would like to ask how big the databases are ?

[loudbutt]

It looks like it’s 182 MB. I don’t even get what it’s trying to read / write though. I didn’t see anything running in the background when I saw the Sqlite busy messages.

[ChuckPa]

There are two databases.

1. com.plexapp.plugins.library.db
2. com.plexapp.plugins.library.blobs.db

These both have corresponding -WAL and -SHM files while PMS is running.

Would you mind sharing a ls -lah of the Databases directory ?
(no personal data there)

[loudbutt]

drwxr-xr-x 1 nobody users 1.2K Jun 24 17:13 ./
drwxr-xr-x 1 nobody users  100 Jun 18 22:46 ../
-rw-r--r-- 1 root   root  4.8K Jun 23 19:13 DBRepair.log
-rw-r--r-- 1 nobody users 1.1M Jun 23 19:13 com.plexapp.plugins.library.blobs.db
-rw-r--r-- 1 nobody users 353K Jun 20 02:05 com.plexapp.plugins.library.blobs.db-2025-06-20
-rw-r--r-- 1 nobody users 353K Jun 23 02:06 com.plexapp.plugins.library.blobs.db-2025-06-23
-rw-r--r-- 1 nobody users  32K Jun 24 14:52 com.plexapp.plugins.library.blobs.db-shm
-rw-r--r-- 1 nobody users    0 Jun 23 19:13 com.plexapp.plugins.library.blobs.db-wal
-rw-r--r-- 1 nobody users 174M Jun 24 19:32 com.plexapp.plugins.library.db
-rw-r--r-- 1 nobody users 115M Jun 20 02:05 com.plexapp.plugins.library.db-2025-06-20
-rw-r--r-- 1 nobody users 189M Jun 23 02:06 com.plexapp.plugins.library.db-2025-06-23
-rw-r--r-- 1 nobody users  32K Jun 24 19:32 com.plexapp.plugins.library.db-shm
-rw-r--r-- 1 nobody users 6.7M Jun 24 19:32 com.plexapp.plugins.library.db-wal
-rw-r--r-- 1 nobody users  83M Jun 24 19:38 tv.plex.providers.epg.cloud-78b9b6e0-c15c-4f9d-b21c-bccdd12ef19b.db
-rw-r--r-- 1 nobody users  32K Jun 24 19:38 tv.plex.providers.epg.cloud-78b9b6e0-c15c-4f9d-b21c-bccdd12ef19b.db-shm
-rw-r--r-- 1 nobody users 1.1M Jun 24 19:38 tv.plex.providers.epg.cloud-78b9b6e0-c15c-4f9d-b21c-bccdd12ef19b.db-wal

[loudbutt]

I have apps that plug into Plex - the “usual” cocktail of:

• Overseerr
• SuggestArr
• ListSync
• Dispatcharr (for IPTV)
• Tautulli
• Kometa

I think the heaviest API calls would come from Kometa if anything, but I’d be surprised to find that it would overwhelm the sqlite db

[ChuckPa]

Simple test time (given one user found a problem today by turning off apps),

Turn them off and retest PMS.

If everything is OK, reintroduce one at a time until the threshold is reached then start swapping off/on to see if it’s one or a combined total.

[loudbutt]

I have it running on Unraid using the official Plex docker image.

I’ve attached some pictures of the CPU usage. There are times where a single core would spike, but only really for a second. The second picture is a representation of what the “typical” state would be.

I don’t know how relevant it is, but Unraid is running as a VM on ESXi. The video card is passed through directly to Unraid. 48 of the 64 threads were assigned to the Unraid VM with 64 GB of ECC RAM.

[loudbutt]

Is the EPG data update super database intensive? I see that it’s downloading ~20,700 airings. I usually have that done only at 3 AM so it doesn’t interrupt anything, but for some reason, it’s been updating for quite a long time now.

I use the Plex built-in EPG guide.

[loudbutt]

Dude…you are a life saver! Thank you!