[1.1.27] Error loading save (decompression failed: input is invalid or incomplete)

Place for things which are bugs but we have no idea how to solve them. Things related to hardware, libraries, strange setups, etc.
Post Reply
Letoric
Manual Inserter
Manual Inserter
Posts: 2
Joined: Fri Mar 19, 2021 11:39 pm
Contact:

[1.1.27] Error loading save (decompression failed: input is invalid or incomplete)

Post by Letoric »

Hello. I'm running a lot of mods, however, I've experienced the issue 3-4 times in the last month that I returned to the game. The most significant action I took between the save that loads, and this one, was used the delete chunks command to reduce save size. Other than that, it was typical play with modded Factorio.

Lua used for deleting chunks:
/c local surface = game.player.surface;
game.player.force.cancel_charting(surface);
local chunk_radius = 12;
for chunk in surface.get_chunks() do
if (chunk.x < -chunk_radius or chunk.x > chunk_radius or chunk.y < -chunk_radius or chunk.y > chunk_radius) then
surface.delete_chunk(chunk)
end
end


Attaching the save that works, and the save that fails, in case you can repro it with deleting chunks. I could not.1Pyanodons2.zip is the working save.

Something of note that a majority of people would not have, is I run "Primo Cache" to boost performance and also save writes to my SSD with the delayed write functionality. I've been using this for at least 6 months without issue, I believe it's been longer.

Edit:
I enabled logging under developer settings, and got this extra tidbit.
128.511 Loading map D:\Factorio\saves\1Pyanodons1.zip: 22484958 bytes.
128.614 Loading level.dat: 80553202 bytes.
128.618 Info Scenario.cpp:197: Map version 1.1.27-0
130.557 Warning Map.cpp:346: Map loading failed: Decompression failed: input is invalid or incomplete.
131.379 Error AppManagerStates.cpp:1590: Decompression failed: input is invalid or incomplete.

2nd Edit:
Issue occurred again on a brand new game, within the first 6 hours of play, and the save from 11 minutes before still works. This one had no chunks deleted. I know you aren't planning on any feature enhancements to the game at this point, but if you could consider adding more rolling for the logs so we could backtrack easier, that would be great for helping us help you to help us.
2Pyanodons1.zip is the broken save
2Pyanodons2.zip is the good save.
Attachments
2Pyanodons2.zip
(16.65 MiB) Downloaded 66 times
2Pyanodons1.zip
(16.82 MiB) Downloaded 66 times
factorio-current.log
(97.6 KiB) Downloaded 68 times
mod-list.json
(9.05 KiB) Downloaded 73 times
1Pyanodons2.zip
(32.05 MiB) Downloaded 92 times
1Pyanodons1.zip
(21.44 MiB) Downloaded 65 times

Rseding91
Factorio Staff
Factorio Staff
Posts: 13178
Joined: Wed Jun 11, 2014 5:23 am
Contact:

Re: [1.1.27] Error loading save (decompression failed: input is invalid or incomplete)

Post by Rseding91 »

Hello, thanks for the report. Looking at the save files with errors and your edits the only conclusion I come to is that you have bad or failing hardware on your end. That specific error means that the save file has been corrupted when writing to disk: something that console commands or running world state wouldn't effect in any way. It's failing to decompress the raw data before it even gets to the loading of the game data.

I recommend running a memory test and if you're overclocking your CPU or RAM to try setting them back to default speeds. Aside from looking into failing hardware there's nothing for us to do on our end; the error getting reported when loading is working correctly and there's nothing we can do about it. It shouldn't be corrupt and something outside of our control is corrupting it.
If you want to get ahold of me I'm almost always on Discord.

Letoric
Manual Inserter
Manual Inserter
Posts: 2
Joined: Fri Mar 19, 2021 11:39 pm
Contact:

Re: [1.1.27] Error loading save (decompression failed: input is invalid or incomplete)

Post by Letoric »

Thank you for the response.
RAM:
I ran a full RAM test of 64GB of RAM overnight, and it passed with 0 errors.

Storage:
The save files were being corrupted on 2 completely separate drives, originally they were on the C drive in appdata, but later I added a directory junction to my SSD, and saw the same issue repeat which in theory, makes the HDD failure very low probability for the cause of failure, since that's 2 separate cables, 2 separate connections on the board, 2 separate power cables.

CPU:
CPU was overclocked, but not heavily, I used to OC to over 4GHz, it was only pushing 3.8GHz for this, however the I7-3930 is well known for it's ability to OC well, so that seems unlikely (though definitely still possible). I've dropped it down to 3.48Ghz (Asus Sabertooth X79 BIOS reset to defaults and then AUTO OC execution process initiated) to see if that proves to solve the issue. Can you possibly investigate more on your side as well, or are you 100% that it can't be on your side?

Post Reply

Return to “1 / 0 magic”