I think this is a pretty good step in the right direction, but I feel it’s a little vague at the moment.
What is CoreMemory:Total?
I feel this is good but it needs some more elaboration/documentation.
I think this is a pretty good step in the right direction, but I feel it’s a little vague at the moment.
What is CoreMemory:Total?
I feel this is good but it needs some more elaboration/documentation.
+1
Is there any documentation that explains why total is so far above the combination of parts? Where could this untracked memory be coming from?
great update i love it thamk u roblo ;33
just wanted to know why my chart goes downwards after the 4th hour
could someone pls explain to me what “system” means and why it takes up so much space
The player amount it mentions here is the server’s peak player count. For example, if you had 50 max players but the highest player count a given server had seen was only 30, that server would get 9.4GB of memory, even if the server had dropped down to 10 players
Good feedback we should ideally have some documentation that defines what these categories are. I’ll look into it - might have to be for 2025.
Great question! (We should add this to the FAQ).
At each point along the X axis, we aggregate the values from all servers that reached that point. There’s a natural decreasing sample size as we move along the X-axis, as fewer and fewer servers live this long. In a game with loads of users, this is probably fine as the sample sizes are large enough to eliminate the noise. The decrease in your chart is likely due to fewer servers being available to aggregate around hour 5 and thus the data point being noisier.
being able to see memory leaks from a glance is a good thing, its a net positive for the platform and development process
Let’s see …
You didn’t think that “question” through - did you ?
It was pretty detailed - and welcomed by 99% of Devs.
What you do “NEED”… is to pay attention to all the multitudes of updates & upgrades they have been doing… than feeling entitled.
They more than welcome any issues you might have… but we see you didn’t name any.
This is a great update Roblox - Thank You - & my sympathies to you for the ranters.
not every developer has the ability to get a large amount of QA testers
for you to even say that is just extremely dumb and disrespectful to nearly every small dev out there.
i myself dont have a team of QA testers for my games, and instead just test with my friends (usually 1-3 at most, and we sometimes spend up to an hour playing if we have a lot of fun) and take notes on what i need to fix.
anyways, this update is also very useful for catching memory leaks, and its a very good feature to see what exactly happens as the server age increases.
This chart might be helpful if I knew what values are ‘bad’ and then what to do about it. Having the System grow 600% doesn’t sound good, but is it bad? How would you go about figuring out what is leaking to cause ‘System’ to grow.
??
this looks like a leak to me but i have no idea what could be leaking
placememory:physicsparts seems to be the big one. Does :destroy not get rid of the parts? I shouldn’t be having parts building up over time. Nothing is visibly lying around the map.
All parts created by players should be ‘cleaned up’ by either :destroy() or fallen parts destroy height
oh! that’s understandable, thanks for your response :3
could you maybe add a mode to switch between the aggregation mode or a way to control at which point it starts, or also add a way to check the worst server’s stats vs the best server’s?
my game used for the previous analytics (Wubby) is a world-building game, meaning that servers don’t usually last for more than a few hours because, frequently, when the world’s owner leaves, everyone else in the world (server) leaves. also, someone can just come and create a weird contraption to crash their world or some worlds might have more content in them than others, which creates a big difference in memory between many servers.
if you could add tools to spot memory leaks in these cases, it would help us a ton, thamk u 4 ur patience and ur response :3
If you didn’t read it, don’t comment on it.
If you did read it, then you are ignoring the benefit of additional insight into memory. It’s an improvement, I see no reason to complain.
600% + team lets gooo
Yeahh I have a lot of work to do…