Best news: do this asap to lower your Cloud Storage (AWS S3) bill

Quick edition to share genuinely good news.

Starting with Nuxeo LTS 2023 HF39, Nuxeo can store blobs in S3 using Intelligent Tiering by default as the storage class. https://connect.nuxeo.com/nuxeo/site/marketplace/package/nuxeo-2023-HF39

If you store a lot of binaries in S3, this is a big cost saver.

As you might recall from my previous edition, this was already possible, but it required configuring an S3 Lifecycle rule to transition objects from STANDARD to INTELLIGENT_TIERING in the bucket. But of course, this was not possible if you did not manage the S3 bucket yourself.

HF39 introduces a new Nuxeo configuration property, nuxeo.s3storage.storageClass, that allows you to specify which storage class should be used to store blobs in an S3 bucket (even if you are deployed in Hyland Cloud and you have no access to the AWS infrastructure). Supported values are STANDARD (default) or INTELLIGENT_TIERING.

So install HF39 and make sure you add:

nuxeo.s3storage.storageClass=INTELLIGENT_TIERING

in nuxeo.conf.

You can also fine tune which storage class should be applied to the full text blobs extracted from binaries when nuxeo.vcs.fulltext.storedInBlob=true, using:

nuxeo.s3storage.fulltext.storeInBlob.storageClass=INTELLIGENT_TIERING

Alternatively, when defining your own S3 blob provider XML contribution, you can also specify the storageClass property:

<extension target="org.nuxeo.ecm.core.blob.BlobManager" point="configuration">
  <blobprovider name="myBlobProvider">
    <class>org.nuxeo.ecm.blob.s3.S3BlobProvider</class>
    ...
    <property name="storageClass">INTELLIGENT_TIERING</property>
  </blobprovider>
</extension>

Important note: the INTELLIGENT_TIERING class can only be used when the underlying bucket is configured to use only immediate (synchronous) access tiers; Nuxeo does not handle Archive Access and Deep Archive Access tiers.

The HF release notes do not specify any impact on existing assets already in the bucket. But as per AWS documentation, you can still use a lifecycle configuration on the bucket to transition existing assets into the Intelligent Tiering class and lower your bill even further (see my previous article):

Per AWS doc: When you add a Lifecycle configuration to a bucket, the configuration rules apply to both existing objects and objects that you add later.

← All insights