When adding a list of movies via drag n drop memory hits the roof
I have found restarting Tixati seems to help but I have lots more to be added

You mean dragging and dropping a lot of torrent files from Explorer to Tixati window, or something else?

yes  dragging and dropping a lot of folders from Explorer to Tixati
The memory stays real high untill I restart Tixati then it drops back to normal

There seems to be a memory leak. After 3 weeks of uptime Tixati had 3.5 GB memory swapped, obviously allocated, but idle. Without zswap this would have blown up my machine.

Same here on windows 11 latest updates

Yeah, I'm constantly getting a "not responding" message on and off. And if I try to kill the process, it takes forever. Plus, when I restart it, it takes hours to check the torrent before it starts to download again.

Have you connected to all of the available channels? Then try reading their descriptions, starting from the biggest one.

It eats all the available memory on windows. And I do not use the channels, just the magnet links and torrent files.

how many torrents do you have running?
do they have many seeders/downloaders?

press the help button, question mark on the right.
go to diagnostics, memory state, all, and run that graph.
then run the graph for the one that takes up your memory.
post those pics here.

maybe run the graphs for memory allocation rate as well.

Answer and advice from Gemini AI.

It sounds like you’ve run into a classic Memory Leak (or at least aggressive caching) that occurs when Tixati’s shell integration has to "handshake" with Windows Explorer for a large volume of files.

When you drag and drop a folder, Tixati doesn't just see a folder name; it triggers a recursive scan of every file inside, generates metadata, calculates hashes, and populates the UI tree. In version 3.42, this process seems to "hoard" RAM that isn't always released until the process is killed.

Here is how to stop the "memory hit" without needing to restart every time:
1. The "Meta-Data" Culprit

When you drag and drop, Tixati tries to load the "metadata" (the file list and structure) for every single movie at once. If you are dropping 50+ folders, Tixati is trying to hold 50+ separate "Add Torrent" dialog states in its active memory.

   The Workaround: Instead of dragging folders directly into the main window, try dropping them into a "Watched Folder."

   Go to Settings > Transfers > Meta-Info.

   Enable a "Watch Folder" and move your .torrent or magnet files there. Tixati processes these sequentially and more efficiently than a bulk drag-and-drop.

2. Disable "Pre-Allocation" Thrashing

If Tixati is set to "Pre-allocate all files" (found in Settings > Files > Local File Maintenance), the moment you drop those folders, Tixati starts asking Windows to reserve GBs of space on your SSD. This creates a massive "I/O Wait" queue that often looks like a memory spike in Task Manager because the "System Cache" is ballooning to keep up with the file creation requests.

   Try this: Change "File Allocation" to "Sparse" or "None" temporarily to see if the RAM usage stabilizes.

3. Clear the "Undo" Buffer

Tixati keeps a history of UI actions. When you bulk-add, it keeps a "snapshot" of those additions in case you want to undo or modify them.

   The Fix: Go to Settings > User Interface.

   Look for any logging or history settings and set them to "Minimal."

4. Check the "Shell Extension"

Sometimes the memory leak isn't Tixati itself, but the Tixati Shell Extension (the part that talks to Windows Explorer).

   Go to Settings > Shell Integration.

   Uncheck the "Drag and Drop" or "Context Menu" options.

   Instead, use the "Add" button inside Tixati and browse to the folders. This uses a different Windows API that is often much lighter on resources than the "Drag and Drop" handler.

Pro Tip for Massive Lists

If you have hundreds of movies to add, don't add them all in "Active" mode. 1.  Drag them in.
2.  In the "Add" dialog, select "Add Paused."
3.  Once they are all in the list and the memory has settled, start them 5 at a time. This prevents the "Hash Check" from hitting your CPU and RAM all at once.