Florian R. Hölzlwimmer

Tensorflow 1.9.0: ``` Error encountered when serializing while_context. Type is unsupported, or the types of the items don't match field type in CollectionDef. ```

Hi, I'd like to try Sled on a python project for parallel writes from multiple processes (very tricky with RocksDB). Is there an update on this issue?

+1 for rclone serve syncthing It would be damn awesome if rclone would integrate with syncthing!

> ``` > #!/bin/bash > > while [[ true ]] ; do > > # performs synchronizations / copy > rclone sync .... > > # waiting for something to...

Hi @ihnorton, thank you for your investigation. Your responsiveness is a real pleasure :) To the serial running: I guess you have to repartition your spark dataframe after creating it....

Is it possible that creating a TileDB Ctx on every opening of the array in parallel could cause the issues? After some trial and error, I now always get `TileDBError:...

I accidentally found that PyTest captured the error log I was missing all the time: `Exception: TileDBError("[TileDB::IO] Error: Cannot lock filelock 'GT/__lock.tdb'; Input/output error")` In other words, our array is...

@stavrospapadopoulos this is not correct: > * If one tiledb process is performing **consolidation** while the other is reading, then there may be issues. This is because tiledb uses filelocks...

fyi, we very successfully run TileDB on both NFS and CephFS :)

Any updates on this? Having a Rust REPL would be THE killer-feature compared to a huge bunch of other compiled languages!