Adding a GIR Object Storage Volume

Once an AWS S3 storage system is registered in Hammerspace, you can add an Object Storage Volume with the GIR storage class. The GIR storage class is supported only on AWS S3; a generic S3 endpoint accepts the Standard storage class and rejects GIR.

The storage class of an OSV is set at creation and cannot be changed afterward. Plan your OSV layout carefully before adding volumes.

Use a new, empty S3 bucket for a GIR OSV. Adding a bucket that already contains data is not a supported configuration, and an existing volume cannot be converted to GIR.

Using the Web UI

Go to Infrastructure Storage Systems and click the + Volume button next to the AWS S3 storage system.
In the Add Volumes wizard, select the S3 bucket or buckets to add as volumes.

On the Data tab, locate the Archive Type field. Leave Archive Type unselected to use Standard storage, or select AWS Glacier Instant Retrieval to create a GIR volume.

When Archive Type is unselected, the volume is added as a regular Standard-class OSV and the Review & Add screen displays a dash for the field. Only volumes explicitly set to AWS Glacier Instant Retrieval are managed as GIR.

Review and adjust the capacity, availability, durability, and IP settings as needed, or accept the defaults.
On the Review & Add screen, confirm that Archive Type shows AWS Glacier Instant Retrieval before submitting. A dash means the volume will be created as Standard.
Click Add Volume.

Using the Admin CLI

object-volume-add \
  --node-name <aws-storage-system-name> \
  --logical-volume-name <bucket-name> \
  --storage-class AWS_GLACIER_INSTANT_RETRIEVAL \
  --create-placement-objectives

The --storage-class parameter accepts the following values (not case-sensitive):

STANDARD — default if omitted; adds a standard-class OSV
AWS_GLACIER_INSTANT_RETRIEVAL — adds a GIR-class OSV

The --create-placement-objectives flag automatically creates place-on, confine-to, and exclude-from objectives for the new OSV. Using this flag is recommended when adding a GIR OSV so the objectives are ready when you configure share archive rules.

--storage-class is rejected with an error on Azure, GCP, and Hammerspace S3 storage nodes. GIR is AWS S3 only.

Default Settings for GIR OSVs

A new GIR OSV receives the same default settings as a standard object storage volume:

Setting	Default
High / Low capacity threshold	75% / 60%
Availability	99.99%
Durability	99.9999999%
Online delay	5 minutes

Setting

Default

High / Low capacity threshold

75% / 60%

Availability

99.99%

Durability

99.9999999%

Online delay

5 minutes

You can adjust these when adding the volume or accept the defaults. Note that although the capacity thresholds are set on GIR OSVs, watermark-based data evacuation is not applied to them (see Mobility behavior).

Verifying the OSV Was Added Correctly

After adding the volume, confirm the configuration with object-volume-list --name, naming the volume as <storage-system-name>::<bucket-name>. (The plain object-volume-list summary does not show the storage class or the volume groups.)

Command:

object-volume-list --name <aws-storage-system-name>::<bucket-name>

In the output, look for the Storage class line and the archive-object-volumes entry under Locations:

Expected output:

ID:                      d4385b5f-e24d-4759-bc62-5b97a1023e49
Name:                    aws-s3::docs-gir-bucket
Internal ID:             268435473
State:                   OK
Node:                    aws-s3
Oper state:              Up
Admin state:             Up
Capabilities:            High threshold:           75%
                         Low threshold:            60%
                         Availability:             99.99%
                         Availability (effective): 99.99%
                         Availability drop:        Disabled
                         Durability:               99.9999999%
                         Durability (effective):   99.9999999%
                         Online delay:             5 minutes
Total capacity:          Unlimited
Physical used:           0B
Logical used:            0B
Physical free:           Unlimited
Naming type:             Hash named
Compression type:        No compression
Chunking type:           Fixed-size
Storage class:           AWS Glacier Instant Retrieval
Encrypt using KMS:       None. No Kms Configured.
Created:                 2026-09-22 13:23:45 UTC
Modified:                2026-09-22 13:23:45 UTC
Locations:
                         [Type: Volume Group, ID: 4bc06571-70ea-4f20-b9df-dfd3d92c7ce6, Name: archive-object-volumes]
                         [Type: Node, ID: fe49a56c-aa46-48ca-b762-1fc5f33d08dc, Name: aws-s3]
                         [Type: Object Storage Volume, ID: d4385b5f-e24d-4759-bc62-5b97a1023e49, Name: aws-s3::docs-gir-bucket]
                         [Type: Volume Group, ID: 275f5acb-f7a3-4010-9127-d5a9f12fca02, Name: all]
Logical volume:          Type:                    Object Store Logical Volume
                         ID:                      fb1d5485-48aa-48c3-9081-48ed007284ac
                         Name:                    docs-gir-bucket
                         Ownership:               Not shared
GC enabled:              Yes

Standard OSVs show Storage class: Standard. If the GIR OSV does not appear in archive-object-volumes, see the troubleshooting section.

To confirm that data uploaded to the OSV is being correctly tagged in S3, use the AWS CLI after placing a file on the volume:

Command:

aws s3api head-object \
  --bucket <bucket-name> \
  --key Hammerspace_v2/<object-path>

The response should include "StorageClass": "GLACIER_IR".