Files bigger than 16MBs not yet supported

[freezing]

Although the exception says:
java.lang.AssertionError: files bigger than 16GB not yet supported;
According to my tests, files bigger than 16MB are not supported.

According to the following lines:
https://github.com/marschall/memoryfilesystem/blob/master/src/main/java/com/github/marschall/memoryfilesystem/MemoryInode.java#L32
https://github.com/marschall/memoryfilesystem/blob/master/src/main/java/com/github/marschall/memoryfilesystem/MemoryInode.java#L356

IndirectBlocks is a matrix of size BLOCK_SIZE * BLOCK_SIZE which is 4k * 4k = 16MB.

Also, I think there should be two constants, NUMBER_OF_BLOCKS * BLOCK_SIZE. It's more readable.

Can it be changed to support a file of size at least a couple of Gigs?
The exception message should be changed as well:
https://github.com/marschall/memoryfilesystem/blob/master/src/main/java/com/github/marschall/memoryfilesystem/MemoryInode.java#L360

[marschall]

Regarding the additional constant and exception message, these are good points and I fixed them in d6f915a.

Now for the maximum file size how big do you really need? Remember this will all be heap memory. The quick fix is to simply up the limits but this will cause really large objects in the heap. The proper solution would be to implement double indirect addressing.

[freezing]

Thanks.

I think 1GB should be more than enough at the moment. I understand that it will be heap memory, but for now the quick fix will work just fine.

For 1GB it should take 300k of blocks to be initialized. Each of them will be empty unless something is written to the file. It should be fine since not all of the files will be that large.

[develar]

Any news?

[marschall]

@develar no new so far, is this a deal breaker for you?

[develar]

Library is widely used in tests, but some tests create large files. So, will be great if limit will be raised (at least, 512 MB).

…

On Sun, 26 Jul 2020, 14:56 Philippe Marschall, ***@***.***> wrote: @develar <https://github.com/develar> no new so far, is this a deal breaker for you? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#71 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACVTXUGB76IZMQFDSTGMDLR5QRZFANCNFSM4B2FIRCQ> .