trilium becoming unsable because stalls and cannot searching without stalling

[whyeverythingsohard]

Description

Hello I've been using trilium for around 7 months now and lately I have not been able to use it. I have all sorts of stuff like health records and doctor appointments because so many health issues I need to keep track of phone numbers with different specialists and insurance departments and people. But now I cannot use the search function because trying to search anything stalls the whole program. what can I do to make it more usuable? is trillium being on a nas a issue?

TriliumNext Version

.97

What operating system are you using?

Windows

What is your setup?

Local (no sync)

Operating System Version

win10

Error logs

No response

[dosubot]

Trilium can stall or become unresponsive during search if your database is very large or contains extremely large notes—this is a known issue, especially if you have notes with very high word counts, as phrase matching can take a long time and freeze the app. There’s no user-side fix for this yet, but maintainers plan to address it in future updates by limiting the note size included in searches (source).

Running Trilium with the database stored on a NAS can also cause performance problems. SQLite (which Trilium uses) is not optimized for networked storage and can introduce latency or even data corruption if accessed directly over a network share. The recommended setup is to keep your database on a local disk, or use Trilium’s server-client sync mechanism if you want to use a NAS (source) (source).

To improve performance:

• Move your Trilium database file to your local drive if it’s currently on a NAS.
• Check your database size (the document.db file in your Trilium data folder). If it’s very large (e.g., over 100MB), consider archiving or splitting up very large notes.
• Run built-in maintenance: use the database vacuum and consistency check features (these are available in the UI or via the API, depending on your setup) to optimize and repair the database (source).
• Always back up your database before running maintenance.

If you still have issues after these steps, try to collect error logs or check the metrics endpoint for database stats, which can help diagnose if your database is unusually large or has problematic notes.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[eliandoran]

@whyeverythingsohard , generally search is a bit demanding if your database is big.

Can you check:

1. How big is the database? You can check this from the backend logs when the server starts or estimate it by selecting the root note -> Note info -> calculate size.
2. What is the CPU and RAM of the server that Trilium is running on?

[whyeverythingsohard]

@whyeverythingsohard , generally search is a bit demanding if your database is big.

Can you check:

1. How big is the database? You can check this from the backend logs when the server starts or estimate it by selecting the root note -> Note info -> calculate size.

2. What is the CPU and RAM of the server that Trilium is running on?

1.the database file is around 3gbs big and if I try to do what you did with calculate size it just instantly crashes trilium haha sorry I can't get the exact number.

2. Intel Celeron J4125 CPU and 20gbs of ram

[eliandoran]

@whyeverythingsohard , yes, unfortunately the CPU is the bottleneck here. I have a similar situation with a Intel Pentium G2120 @ 3.10GHz.

Since the search is a pretty CPU-intensive operation, there's not much we can do. Perhaps #6839 will increase the speed, since it uses native SQLite operations.

The only thing that can be done on your side is to try to make searches as specific as possible by setting an ancestor, or enabling Fast search (which will disable search in content).

[whyeverythingsohard]

@whyeverythingsohard , yes, unfortunately the CPU is the bottleneck here. I have a similar situation with a Intel Pentium G2120 @ 3.10GHz.

Since the search is a pretty CPU-intensive operation, there's not much we can do. Perhaps #6839 will increase the speed, since it uses native SQLite operations.

The only thing that can be done on your side is to try to make searches as specific as possible by setting an ancestor, or enabling Fast search (which will disable search in content).

Thank you Eliandoran appreciate you educating me