This change improves the performance of writing to sockets with the
default Java URL connection HTTP client, by enlarging the buffer used
for socket writes from an implicit hard-coded 4KB / 8KB buffer to a
configurable 32KB buffer.
The buffer size is now controlled by the following property with the
following default value:
jclouds.output-socket-buffer-size: 32768
The implementation is based on a variant of ByteStreams.copy (written as
ByteStreams2.copy) which accepts the buffer size as an argument, unlike
the original Guava code that uses a hard-coded size.
The change was done directly within the loop that copies the input
stream to the output stream, and not by wrapping a BufferedOutputStream
around the existing output stream, in order to avoid copying the payload
twice.
On some platforms this change can improve both the putBlob throughput
and the total CPU consumption.
* JCLOUDS-1496: Update maven-compiler-plugin for increased JDK compatibility
* increase maven-compiler-plugin version
* A space change to trigger jenkins again
* Another space change to trigger jenkins again
These should provide a descriptive second argument, not the same as
the first argument which is null in the failure case. This also found
a logic error in CreateVolumeResponseHandler.
Previously getBlobKeysInsideContainer returned all keys and filtered
in LocalBlobStore. Now getBlobKeysInsideContainer filters via prefix
which can dramatically decrease the number of keys returned,
especially for the filesystem provider. Further optimizations are
possible for delimiter.
Previously jclouds attempted to slice non-repeatable InputStream
Payloads in order to upload sequentially. This never worked due to
mutating the single stream via skip and close. Also backfill test
which spuriously succeeded.
Changes order in parameters to keep properties as last parameter
Removes wrong @Nullable parameters
Moves LoadBalancer and PublicAddress to new package with SKU as inner
class
Moves LoadBalancerSKU inside LoadBalancer. Creates new loadbalancer package
Refactors PublicAddress domain objects
Extracts SKU from LoadBalancer
LoadBalancer SKU as inner class
PublicIPAddress SKU as inner class
Reorder parameters sku and properties
B2 requires two parts to use multi-part upload but jclouds will
attempt to use one with small uploads. Instead it should fall back to
single-part when the content length is small.
Prior commit introduced a bug in the computation of the MPU ETag value,
where it was concatenating strings, rather than operating on the bytes
of the integer value.
S3 uses a different ETag for multipart uploads (MPUs) than regular
objects. The ETag consists of the md5 hash of the concatenated ETags of
individual parts followed by the number of parts (separated by "-").
The patch changes the LocalBlobStore's implementation of
CompleteMultipartUpload to set the S3-style ETag before calling
putBlob() and return that ETag to the caller.
To simplify testing, a new protected method with a default NOOP
implementation is added to the BaseBlobIntegrationTest. It allows
providers to further verify MPUs (i.e. ensuring the correct ETag has
been stored alongside the object).
Olaf Flebbe
Dori Polotsky
Andrew Gaul
Dani Estevez
d065488
Jean-Baptiste Onofré
Joe Meiring
duc
Simone Locci
John McDonnell
kraza
Timur Alperovich
filecatalyst
Trent Schmidt
Ignasi Barrera