Add storage container docs (#294)

* Add storage container docs Signed-off-by: greenmoon55 <greenmoon55@gmail.com> * Update storage container doc Signed-off-by: greenmoon55 <greenmoon55@gmail.com> * Move doc to Model Storage section Signed-off-by: greenmoon55 <greenmoon55@gmail.com> * Update docs/modelserving/storage/storagecontainers.md Signed-off-by: Dan Sun <dsun20@bloomberg.net> * Add warning section Signed-off-by: Dan Sun <dsun20@bloomberg.net> * Update storagecontainers.md Signed-off-by: Dan Sun <dsun20@bloomberg.net> --------- Signed-off-by: greenmoon55 <greenmoon55@gmail.com> Signed-off-by: Dan Sun <dsun20@bloomberg.net> Co-authored-by: Dan Sun <dsun20@bloomberg.net>
kserve · Oct 14, 2023 · 74693a7 · 74693a7
1 parent abdc5bd
commit 74693a7
Show file tree

Hide file tree

Showing 2 changed files with 65 additions and 0 deletions.
diff --git a/docs/modelserving/storage/storagecontainers.md b/docs/modelserving/storage/storagecontainers.md
@@ -0,0 +1,64 @@
+# Storage Containers
+
+KServe downloads models using a storage initializer (initContainer). For example, this is [the default storage initializer implementation](https://github.com/kserve/kserve/blob/79f2a48d0c9c72b034127170e38d6b29b927f03a/python/kserve/kserve/storage/storage.py). KServe introduced **ClusterStorageContainer** CRD in 0.11 which allows users to specify a custom container spec for a list of supported URI formats. 
+
+A _ClusterStorageContainer_ defines the container spec for one or more storage URI formats. Here is an example of a ClusterStorageContainer that corresponds to the default storage initializer. Note that this is incluced in the [helm chart](https://github.com/kserve/kserve/blob/79f2a48d0c9c72b034127170e38d6b29b927f03a/charts/kserve-resources/templates/clusterstoragecontainer.yaml). 
+
+```yaml
+apiVersion: "serving.kserve.io/v1alpha1"
+kind: ClusterStorageContainer
+metadata:
+  name: default
+spec:
+  container:
+    name: storage-initializer
+    image: kserve/storage-initializer:latest
+    resources:
+      requests:
+        memory: 100Mi
+        cpu: 100m
+      limits:
+        memory: 1Gi
+        cpu: "1"
+  supportedUriFormats:
+    - prefix: gs://
+    - prefix: s3://
+    - prefix: hdfs://
+    - prefix: webhdfs://
+    - regex: "https://(.+?).blob.core.windows.net/(.+)"
+    - regex: "https://(.+?).file.core.windows.net/(.+)"
+    - regex: "https?://(.+)/(.+)"
+```
+
+In a _ClusterStorageContainer_ spec, you can specify container resource requests and limits, and a list of supported URI formats that this image supports. KServe can match the URI either with `prefix` or `regex` .
+
+!!! warning
+
+    If a storage URI is supported by two or more _ClusterStorageContainer_ CRs, there is no guarantee which one will be used. **Please make sure that the URI format is only supported by one ClusterStorageContainer CR**.
+
+
+If you would like to use a custom protocol `model-registry://`, for example, you can create a custom image and add a new ClusterStorageContainer CR like this:
+
+```yaml
+apiVersion: "serving.kserve.io/v1alpha1"
+kind: ClusterStorageContainer
+metadata:
+  name: abc
+spec:
+  container:
+    name: storage-initializer
+    image: abc/custom-storage-initializer:latest
+    resources:
+      requests:
+        memory: 100Mi
+        cpu: 100m
+      limits:
+        memory: 1Gi
+        cpu: "1"
+  supportedUriFormats:
+    - prefix: model-registry://
+```
+
+## Spec Attributes
+
+Spec attributes are in [API Reference](/website/reference/api/#serving.kserve.io/v1alpha1.ClusterStorageContainer) doc.
diff --git a/mkdocs.yml b/mkdocs.yml
@@ -52,6 +52,7 @@ nav:
                 - Concept: modelserving/inference_graph/README.md
                 - Image classification inference graph: modelserving/inference_graph/image_pipeline/README.md
           - Model Storage:
+                - Storage Containers: modelserving/storage/storagecontainers.md
                 - Azure: modelserving/storage/azure/azure.md
                 - PVC: modelserving/storage/pvc/pvc.md
                 - S3: modelserving/storage/s3/s3.md