-
Notifications
You must be signed in to change notification settings - Fork 454
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
High amount of Disk write requests in 8.0.4 #8272
Comments
Thanks for reporting! That is not intended. Our database (ORM library) should take care of batching writes to your disk to avoid exactly this issue. Perhaps we misconfigured it? We'll check it out. |
Thnx for diving into this! Wow, 24.204 write req/s average 😱😨😱 @qstokkink Please try to reproduce this in the Tribler lab. Hopefully this can be fixed 💪 |
@egbertbouman mitigated the disk writes in #8274 However, my earlier claim..
..shockingly, did not turn out to be true. We will need to make some bigger changes in our own code base to properly solve this issue. Because this issue requires a big refactoring effort, we'll schedule this for the next stable release, |
very very high disk usage still persists in version 8.0.7, hanging (freezing) the computer. especially, for torrents above 5 GB. |
Our first step toward solving this will be #8402. After that has been merged, we'll reevaluate the disk usage. |
After #8402, we still have quite a bit of non-torrent disk I/O. This seems to be due to the following two features:
We should evaluate how we can lower the disk writes of these features. |
I found that tribler makes very high amounts of write requests to the disk, sometimes around 400-600KB/s constantly even when all Downloads are stopped this is going on.
Upon further investigation with Procmon i found that Tribler is doing thousands of WriteFile Requests per second inside.
C:\Users\clusterfuck\AppData\Roaming.Tribler\8.0\sqlite
According to Perfmon around 24.000+ write req/s average:
This may wear out some SSDs on the long term, cause higher CPU usage and maybe it is even slowing down the tribler process itself.
So perhaps it would be usefull if someone has a closer look on what is going on there.
The text was updated successfully, but these errors were encountered: