Closed edz504 closed 8 years ago
any chance you can test without batching? and any chance you can check memory use while running (with batching)?
— wouter
sent from my phone. please forgive my tpyos.
On February 16, 2016 4:53:40 PM GMT+01:00, Eddie Zhou notifications@github.com wrote:
Hi there,
I'm trying to open an existing LevelDB (A) with plyvel, create a new LevelDB (B), and write the first X values of A to B. I'm doing this in batch writes of 100, but for some reason, the first few batches write extremely quickly, and then as the process continues, the writing slows down tremendously. I'm talking from < 1 second for the first batch write to ~10 seconds for the 70th batch.
I'd like to know if this is expected (as B grows in size?), or if there's possibly a small mistake I'm making that's slowing the process down.
Also, if anyone has any suggestions for a better way to subset part of A into a new LevelDB, please let me know.
Thanks!
Reply to this email directly or view it on GitHub: https://github.com/wbolster/plyvel/issues/45
Hi -- figured out my problem, it wasn't plylevel related. Thanks, though!
btw keep in mind that write batches are intended for atomicity, and will generally not make writes faster.
Hi there,
I'm trying to open an existing LevelDB (A) with plyvel, create a new LevelDB (B), and write the first X values of A to B. I'm doing this in batch writes of 100, but for some reason, the first few batches write extremely quickly, and then as the process continues, the writing slows down tremendously. I'm talking from < 1 second for the first batch write to ~10 seconds for the 70th batch.
I'd like to know if this is expected (as B grows in size?), or if there's possibly a small mistake I'm making that's slowing the process down.
Also, if anyone has any suggestions for a better way to subset part of A into a new LevelDB, please let me know.
Thanks!