Sometimes on aws-ec2, the client returns null (i.e. node not found).
This fix will cause us retry with the correct node-id, rather than
"forgetting" the original nodeId and retrying with null each time.
* 'master' of github.com:jclouds/jclouds:
Issue 997:support AWS IAM api
added paginated set functionality
glesys: updating to API version 0.1.29
glesys: removing ArchiveDetails (duplicate of Archive)
glesys: bean cleaning, using ConstructorProperties for deserialization (note no serialization annotations as we don't currently serialize any of these beans)
glesys: adjusting ParserModule - removing support for odd dates (no longer present in api) and adding support for "yes"/"no" Boolean fields
Specifically adding the osgi import org.apache.commons.io.input;version=[1.4,3)
Upgrading commons-io dependency of sshj driver to 2.0
Improve Maven build time dramatically by moving source and javadoc to profiles.
Issue 992: Fix AWS S3 to work with non-DNS, but still valid, named buckets.
Issue 995: handle illegal argument in vcloud images
Issue 994: fix vcloud sax parsing of namespaces
cloudstack: don't set networkId to default when iptonetworklist is populated
Fix for iptonetworklist (was incorrectly sending ipnetworklist)
Adjusting general strategy to accept @Named in place of @SerializedName. Adding TypeAdapterFactory to handle deserialization based on constructor annotations (Inject/Named and/or ConstructorProperties).
added min-disk to compute2.clj
Issue 988:Extra port added to swift url
consistent ordering of hardware
fixed missing provider name on ninefold test
Eliminate unlikely transient blobstore TOCTOU bug
This makes a *huge* difference in build time (mvn clean install
-DskipTests takes about 1/3 as long with this in place). It moves
source and javadoc into profiles, so that if you want to build the
source jars and javadoc jars, you specify -Psrc and -Pdoc respectively.
Prior to this commit, jclouds wouldn't allow you to interact with any buckets
in S3 that were named with uppercase characters. Per AWS S3 docs, this
non-standard naming is valid in US regions only. This update fixes jclouds so
that it can interact with, and even attempt to create, buckets with uppercase
characters for AWS S3 without actually impacting other S3 implementations. This
fix also will not have any impact in non-US regions other than instead of a
bucket name validation error you'll get an InvalidBucketName error back from
AWS S3 when you attempt to create a bucket with an uppercase character in a
non-US region. To summarize, nothing changes other than US regions now allow
creation of bucket names with upper case characters and jclouds now can
interact with these non-standard named buckets without failure.
Adrian Cole
h h
Adam Lowe
Aled Sage
David Ribeiro Alves
Nirmal
Andrew Bayer
Jeremy Whitlock
Andrei Savu
Andrew Gaul