Timothy N. Miller
Timothy N. Miller
@parthchandra Would you mind having a look at my I/O performance optimization plan for ParquetMR? I think we should coordinate, since we have some ideas that might overlap what we...
@parthchandra One thing that confuses me a bit is that these buffers have only ByteBuffer inside them. There's no actual I/O, so it's not possible to block. Do you have...
Is byte (and arrays and buffers of bytes) the only datatype you support? My PR is optimizing code paths that pull ints, longs, and other sizes out of the data...
This is interesting, because when I did profiling of Trino, I found that although I/O (from S3, over the network no less) was significant, even more time was spent in...
> > Is byte (and arrays and buffers of bytes) the only datatype you support? My PR is optimizing code paths that pull ints, longs, and other sizes out of...
That batch reader in Presto reminds me of some of the experimental changes I made in Trino. I modified PrimitiveColumnReader to work out how many of each data item it...
Alright. You have a point. If the maintainers want me to delete that stuff, they can let me know, and I'll go ahead and do it.
> If we change it to be per thread, then would it be a problem in the scenario where short living threads come and go? When the thread stopped, we...
> My question is when a thread exits, we don't have a corresponding evict operation on the map. Using thread pool might be OK if the thread object is not...
I added a test.