add instruction for using alluxiocommon

README.md

@@ -156,6 +156,24 @@ ds2.count()
 # End of Python example
 ```
 
+#### Enable alluxiocommon enhancement module
+
+alluxiocommon package is a native enhancement module for alluxiofs based on PyO3 rust bindings.
+Currently it enhances big reads (multi-page reads from alluxio) by issuing multi-threaded requests to alluxio.
+
+to enable it, first install alluxiocommon package:
+```
+pip install alluxiocommon
+```
+and when start the Alluxio fsspec instance, add an additional option flag:
+```
+alluxio_options = {"alluxio.common.extension.enable" : "True"}
+alluxio_fs = fsspec.filesystem(
+  "alluxiofs", etcd_hosts="localhost", target_protocol="s3",
+  options=alluxio_options
+)
+```
+
 ### Running examples with Pyarrow
 
 ```

alluxiofs/client/core.py

@@ -205,8 +205,8 @@ def __init__(
                 ALLUXIO_COMMON_EXTENSION_ENABLE in options
                 and options[ALLUXIO_COMMON_EXTENSION_ENABLE].lower() == "true"
             ):
-                print("Using alluxiocommon extension..")
-                logger.debug("alluxiocommon extension enabled.")
+                print("alluxiocommon extension enabled.")
+                logger.info("alluxiocommon extension enabled.")
                 ondemand_pool_disabled = (
                     ALLUXIO_COMMON_ONDEMANDPOOL_DISABLE in options
                     and options[ALLUXIO_COMMON_ONDEMANDPOOL_DISABLE].lower()

alluxiofs/core.py

@@ -339,9 +339,10 @@ def tail(self, path, *args, **kwargs):
     def expand_path(self, path, *args, **kwargs):
         raise NotImplementedError
 
-    @fallback_handler
-    def find(self, path, *args, **kwargs):
-        raise NotImplementedError
+    # Comment it out as s3fs will return folder as well.
+    # @fallback_handler
+    # def find(self, path, *args, **kwargs):
+    #     raise NotImplementedError
 
     @fallback_handler
     def mv(self, path1, path2, *args, **kwargs):

benchmark/ray_single_node_image_loader_microbenchmark.py

@@ -447,7 +447,8 @@ def setup_alluxio(args):
         alluxio_options["alluxio.cluster.name"] = args.alluxio_cluster_name
     if alluxio_options:
         alluxio_kwargs["options"] = alluxio_options
-    return fsspec.filesystem("alluxio", **alluxio_kwargs)
+    # alluxio_kwargs["test_options"] = {"log_level" : "DEBUG"}
+    return fsspec.filesystem("alluxiofs", **alluxio_kwargs)
 
 
 if __name__ == "__main__":