Memory Leak - JobTempAlloc

After playing for a few hours, RAM usage had increased to 8GB.

There’s a generated log file in .\Documents\my games\Farthest Frontier\Logs indicating a leak related to Job allocations, likely why its more prevalent after the town size has increased. Below is a snippet of the first few lines:

[WARNING] Internal: JobTempAlloc has allocations that are more than the maximum lifespan of 4 frames old - this is not allowed and likely a leak
[WARNING] To Debug, run app with -diag-job-temp-memory-leak-validation cmd line argument. This will output the callstacks of the leaked allocations.
[WARNING] Property (_InstanceBlend) exceeds previous array size (95 vs 46). Cap to previous size. Restart Unity to recreate the arrays.
[WARNING] Property (_InstanceBlend) exceeds previous array size (88 vs 45). Cap to previous size. Restart Unity to recreate the arrays.
[WARNING] Property (_InstanceBlend) exceeds previous array size (105 vs 59). Cap to previous size. Restart Unity to recreate the arrays.

Those log messages are not related to a memory leak.

I believe there is an issue of memory allocation not fully cleaning up between reloads/returning to menu, which can result in elevated RAM usage. It’s something we’re going to be looking into as we polish the game for v1.0.

How much ram was the game using when you first loaded the town?

Freshly loaded, around 5.5GB.

It’s the same type of behavior i’ve had with similar genre’s of games where having the save loaded for a few hours straight ballons the RAM usage and a reload clears it out usually giving back a little FPS in the process. Rimworld was notorious for a Memory Leak of this nature for a long time.

Even 5.5GB seems high for a town with 200 pops. With the minimum spec listed at 8gb, i would imagine by the time you get to tier 4, the ram usage would exceed that.

i have massive problems with lag as pop gets to 150 pops ish is this why?

It may contribute, but not the sole reason. All city builder/mgmt style games struggle eventually. AI path finding is hugely taxing in these types of games.

But if there is a memory leak, fixing it can help prolong when the “lag” starts happening to get to higher pop counts.

What if we turn off the autosave?

New game, autosave off, only T2 but closing on T3, no lag or buffering.

I’m not referring to lag as in skips/freezing behavior. I’m referring to FPS slowdown in mid to late tier. It happens to all games in this genre, a memory leak could just cause it to happen faster/sooner.

I’m at late tier 3 almost at tier 4 and FPS is still playable, but it hovers around 50+. I do play at 1440p and could lower the resolution, but i should be able to get to tier 4 end game and still have 60+ fps even at 1440p.

There is definetly something wrong with game slowdown - for me, the longer it is being played, the slower it gets. Sometimes I restart the game, just to “restore” some FPS. Reloading a game right after Raids almost doubles my FPS (!).

Any way to fix this problem, I have same issue

My save can not load after I play it in v092 and save it 1 time. So annoy

[WARNING] Internal: JobTempAlloc has allocations that are more than the maximum lifespan of 4 frames old - this is not allowed and likely a leak
[WARNING] To Debug, run app with -diag-job-temp-memory-leak-validation cmd line argument. This will output the callstacks of the leaked allocations.
[WARNING] Internal: JobTempAlloc has allocations that are more than the maximum lifespan of 4 frames old - this is not allowed and likely a leak
[WARNING] To Debug, run app with -diag-job-temp-memory-leak-validation cmd line argument. This will output the callstacks of the leaked allocations.
[WARNING] Internal: JobTempAlloc has allocations that are more than the maximum lifespan of 4 frames old - this is not allowed and likely a leak
[WARNING] To Debug, run app with -diag-job-temp-memory-leak-validation cmd line argument. This will output the callstacks of the leaked allocations.
[WARNING] Internal: JobTempAlloc has allocations that are more than the maximum lifespan of 4 frames old - this is not allowed and likely a leak
[WARNING] To Debug, run app with -diag-job-temp-memory-leak-validation cmd line argument. This will output the callstacks of the leaked allocations.
I confirm that seem like v092abcd all of version cannot fix leak memory when load save file.
I used the same save file to play in v091 and no issue happen. Only v092 above
Please fix it soon