> @@ -158,9 +158,8 @@ private Blob createUpdatedCopyOfBlobInContainer(String
> containerName, Blob in) {
> try {
> if (payload == null || !(payload instanceof ByteArrayPayload)) {
> MutableContentMetadata oldMd =
> in.getPayload().getContentMetadata();
> - ByteArrayOutputStream out = new ByteArrayOutputStream();
> - in.getPayload().writeTo(out);
> - payload = (ByteArrayPayload)
> Payloads.calculateMD5(Payloads.newPayload(out.toByteArray()));
> + byte[] out = ByteStreams.toByteArray(in.getPayload());
> + payload = (ByteArrayPayload)
> Payloads.calculateMD5(Payloads.newPayload(out));
This has the same memory consumption as before. A subsequent commit will
leverage Guava 16 HashingInputStream to calculate the MD5 hash and store the
byte array to disk for the filesystem blobstore and in-memory for transient
while minimizing memory use.
---
Reply to this email directly or view it on GitHub:
https://github.com/jclouds/jclouds/pull/301/files#r9968606
