How to clear cache in esa_snappy Snap 10.0

[saqib.zia]

Hi esa snappy is consuming a lot of cache and RAM how can I prevent this:

To clear RAM I am using:
product.dispose()

but how to clear cache?

[Marco_EOM]

I think you will find my answer to this post helpful.

Why SNAP doesn't clear memory after closing the products?

Java manages the allocated memory. SNAP can only mark the memory as not used anymore. Which it does (in most cases). You can enable a little performance widget in SNAP. Click on the toolbar and select Performance. You will now see a memory graph, showing the used and the allocated memory values. [image] Clicking on the graph tells Java/SNAP to free used memory immediately. The first number, the used memory, will get smaller but the second, the allocated memory, stays almost the same. Window…

You can try to call the garbage collection directly.
See here:

How to free java memory / Snappy

You say you start several times. So freeing the JVM memory will probably not help. Because with every run a new JVM is started, with new memory. But you can try to invoke the garbage collection of the JVM. This would look like the following: import jpy System = jpy.get_type(‘java.lang.System’) System.gc() When using readPixels you should consider doing your computation linewise or in tiles and not the whole band at once.

---

Marco from EOMasters - Mastering Earth Observation -

[saqib.zia]

@How to free java memory / Snappy

jpy.get_type(‘java.lang.System’)

I am already doing this:
jpy.get_type(“java.lang.System”).gc()
still it consumes too much memory

[Marco_EOM]

What do you mean by too much? How much does it consume and how much RAM do you have?

[saqib.zia]

I have 32 gb RAM and I am doing time series calculations.Before starting of script my RAM consumption is 10GB then at the end it jumps to >30 GB
This is my flow:
1.open a product
2.upscale to 10 m
3.In for loop I am cropping polygons and performing indices calculations and dispose all the variables of polygon,ndvi,ndmi etc
4.After for loop I am disposing upscaled_product and product
5.Then I use jpy.get_type(“java.lang.System”).gc()
but in the end my RAM consumption reaches 30 gb

[Marco_EOM]

But you don’t get an OutOfMemory Exception. So, your process works.
Disposing objects and calling the garbage collector are only hints to the Java VM. It can, but don’t have to free memory. But if the memory is needed by another application it is released. But till then it keeps the memory allocated because allocating memory is an expensive operation.

How do you calculate the indices. Do you use GPF or do you calculate them manually?
If you do it manually, it would be good to do it in chunks of multiples lines or tile-wise.

In this example it is done line-wise:

senbox-org/esa-snappy/blob/bff8b40afd47eec61ca3d0398120584bce1fdc76/src/main/resources/esa_snappy/examples/snappy_ndvi.py#L67