1# OCI Image Configuration 2 3An OCI *Image* is an ordered collection of root filesystem changes and the corresponding execution parameters for use within a container runtime. 4This specification outlines the JSON format describing images for use with a container runtime and execution tool and its relationship to filesystem changesets, described in [Layers](layer.md). 5 6This section defines the `application/vnd.oci.image.config.v1+json` [media type](media-types.md). 7 8## Terminology 9 10This specification uses the following terms: 11 12### [Layer](layer.md) 13 14* Image filesystems are composed of *layers*. 15* Each layer represents a set of filesystem changes in a tar-based [layer format](layer.md), recording files to be added, changed, or deleted relative to its parent layer. 16* Layers do not have configuration metadata such as environment variables or default arguments - these are properties of the image as a whole rather than any particular layer. 17* Using a layer-based or union filesystem such as AUFS, or by computing the diff from filesystem snapshots, the filesystem changeset can be used to present a series of image layers as if they were one cohesive filesystem. 18 19### Image JSON 20 21* Each image has an associated JSON structure which describes some basic information about the image such as date created, author, as well as execution/runtime configuration like its entrypoint, default arguments, networking, and volumes. 22* The JSON structure also references a cryptographic hash of each layer used by the image, and provides history information for those layers. 23* This JSON is considered to be immutable, because changing it would change the computed [ImageID](#imageid). 24* Changing it means creating a new derived image, instead of changing the existing image. 25 26### Layer DiffID 27 28A layer DiffID is the digest over the layer's uncompressed tar archive and serialized in the descriptor digest format, e.g., `sha256:a9561eb1b190625c9adb5a9513e72c4dedafc1cb2d4c5236c9a6957ec7dfd5a9`. 29Layers SHOULD be packed and unpacked reproducibly to avoid changing the layer DiffID, for example by using [tar-split][] to save the tar headers. 30 31NOTE: Do not confuse DiffIDs with [layer digests](manifest.md#image-manifest-property-descriptions), often referenced in the manifest, which are digests over compressed or uncompressed content. 32 33### Layer ChainID 34 35For convenience, it is sometimes useful to refer to a stack of layers with a single identifier. 36While a layer's `DiffID` identifies a single changeset, the `ChainID` identifies the subsequent application of those changesets. 37This ensures that we have handles referring to both the layer itself, as well as the result of the application of a series of changesets. 38Use in combination with `rootfs.diff_ids` while applying layers to a root filesystem to uniquely and safely identify the result. 39 40#### Definition 41 42The `ChainID` of an applied set of layers is defined with the following recursion: 43 44``` 45ChainID(L₀) = DiffID(L₀) 46ChainID(L₀|...|Lₙ₋₁|Lₙ) = Digest(ChainID(L₀|...|Lₙ₋₁) + " " + DiffID(Lₙ)) 47``` 48 49For this, we define the binary `|` operation to be the result of applying the right operand to the left operand. 50For example, given base layer `A` and a changeset `B`, we refer to the result of applying `B` to `A` as `A|B`. 51 52Above, we define the `ChainID` for a single layer (`L₀`) as equivalent to the `DiffID` for that layer. 53Otherwise, the `ChainID` for a set of applied layers (`L₀|...|Lₙ₋₁|Lₙ`) is defined as the recursion `Digest(ChainID(L₀|...|Lₙ₋₁) + " " + DiffID(Lₙ))`. 54 55#### Explanation 56 57Let's say we have layers A, B, C, ordered from bottom to top, where A is the base and C is the top. 58Defining `|` as a binary application operator, the root filesystem may be `A|B|C`. 59While it is implied that `C` is only useful when applied to `A|B`, the identifier `C` is insufficient to identify this result, as we'd have the equality `C = A|B|C`, which isn't true. 60 61The main issue is when we have two definitions of `C`, `C = C` and `C = A|B|C`. 62If this is true (with some handwaving), `C = x|C` where `x = any application`. 63This means that if an attacker can define `x`, relying on `C` provides no guarantee that the layers were applied in any order. 64 65The `ChainID` addresses this problem by being defined as a compound hash. 66__We differentiate the changeset `C`, from the order-dependent application `A|B|C` by saying that the resulting rootfs is identified by ChainID(A|B|C), which can be calculated by `ImageConfig.rootfs`.__ 67 68Let's expand the definition of `ChainID(A|B|C)` to explore its internal structure: 69 70``` 71ChainID(A) = DiffID(A) 72ChainID(A|B) = Digest(ChainID(A) + " " + DiffID(B)) 73ChainID(A|B|C) = Digest(ChainID(A|B) + " " + DiffID(C)) 74``` 75 76We can replace each definition and reduce to a single equality: 77 78``` 79ChainID(A|B|C) = Digest(Digest(DiffID(A) + " " + DiffID(B)) + " " + DiffID(C)) 80``` 81 82Hopefully, the above is illustrative of the _actual_ contents of the `ChainID`. 83Most importantly, we can easily see that `ChainID(C) != ChainID(A|B|C)`, otherwise, `ChainID(C) = DiffID(C)`, which is the base case, could not be true. 84 85### ImageID 86 87Each image's ID is given by the SHA256 hash of its [configuration JSON](#image-json). 88It is represented as a hexadecimal encoding of 256 bits, e.g., `sha256:a9561eb1b190625c9adb5a9513e72c4dedafc1cb2d4c5236c9a6957ec7dfd5a9`. 89Since the [configuration JSON](#image-json) that gets hashed references hashes of each layer in the image, this formulation of the ImageID makes images content-addressable. 90 91## Properties 92 93Note: Any OPTIONAL field MAY also be set to null, which is equivalent to being absent. 94 95- **created** *string*, OPTIONAL 96 97 An combined date and time at which the image was created, formatted as defined by [RFC 3339, section 5.6][rfc3339-s5.6]. 98 99- **author** *string*, OPTIONAL 100 101 Gives the name and/or email address of the person or entity which created and is responsible for maintaining the image. 102 103- **architecture** *string*, REQUIRED 104 105 The CPU architecture which the binaries in this image are built to run on. 106 Configurations SHOULD use, and implementations SHOULD understand, values listed in the Go Language document for [`GOARCH`][go-environment]. 107 108- **os** *string*, REQUIRED 109 110 The name of the operating system which the image is built to run on. 111 Configurations SHOULD use, and implementations SHOULD understand, values listed in the Go Language document for [`GOOS`][go-environment]. 112 113- **config** *object*, OPTIONAL 114 115 The execution parameters which SHOULD be used as a base when running a container using the image. 116 This field can be `null`, in which case any execution parameters should be specified at creation of the container. 117 118 - **User** *string*, OPTIONAL 119 120 The username or UID which is a platform-specific structure that allows specific control over which user the process run as. 121 This acts as a default value to use when the value is not specified when creating a container. 122 For Linux based systems, all of the following are valid: `user`, `uid`, `user:group`, `uid:gid`, `uid:group`, `user:gid`. 123 If `group`/`gid` is not specified, the default group and supplementary groups of the given `user`/`uid` in `/etc/passwd` from the container are applied. 124 125 - **ExposedPorts** *object*, OPTIONAL 126 127 A set of ports to expose from a container running this image. 128 Its keys can be in the format of: 129`port/tcp`, `port/udp`, `port` with the default protocol being `tcp` if not specified. 130 These values act as defaults and are merged with any specified when creating a container. 131 **NOTE:** This JSON structure value is unusual because it is a direct JSON serialization of the Go type `map[string]struct{}` and is represented in JSON as an object mapping its keys to an empty object. 132 133 - **Env** *array of strings*, OPTIONAL 134 135 Entries are in the format of `VARNAME=VARVALUE`. 136 These values act as defaults and are merged with any specified when creating a container. 137 138 - **Entrypoint** *array of strings*, OPTIONAL 139 140 A list of arguments to use as the command to execute when the container starts. 141 These values act as defaults and may be replaced by an entrypoint specified when creating a container. 142 143 - **Cmd** *array of strings*, OPTIONAL 144 145 Default arguments to the entrypoint of the container. 146 These values act as defaults and may be replaced by any specified when creating a container. 147 If an `Entrypoint` value is not specified, then the first entry of the `Cmd` array SHOULD be interpreted as the executable to run. 148 149 - **Volumes** *object*, OPTIONAL 150 151 A set of directories describing where the process is likely write data specific to a container instance. 152 **NOTE:** This JSON structure value is unusual because it is a direct JSON serialization of the Go type `map[string]struct{}` and is represented in JSON as an object mapping its keys to an empty object. 153 154 - **WorkingDir** *string*, OPTIONAL 155 156 Sets the current working directory of the entrypoint process in the container. 157 This value acts as a default and may be replaced by a working directory specified when creating a container. 158 159 - **Labels** *object*, OPTIONAL 160 161 The field contains arbitrary metadata for the container. 162 This property MUST use the [annotation rules](annotations.md#rules). 163 164 - **StopSignal** *string*, OPTIONAL 165 166 The field contains the system call signal that will be sent to the container to exit. The signal can be a signal name in the format `SIGNAME`, for instance `SIGKILL` or `SIGRTMIN+3`. 167 168- **rootfs** *object*, REQUIRED 169 170 The rootfs key references the layer content addresses used by the image. 171 This makes the image config hash depend on the filesystem hash. 172 173 - **type** *string*, REQUIRED 174 175 MUST be set to `layers`. 176 Implementations MUST generate an error if they encounter a unknown value while verifying or unpacking an image. 177 178 - **diff_ids** *array of strings*, REQUIRED 179 180 An array of layer content hashes (`DiffIDs`), in order from first to last. 181 182- **history** *array of objects*, OPTIONAL 183 184 Describes the history of each layer. 185 The array is ordered from first to last. 186 The object has the following fields: 187 188 - **created** *string*, OPTIONAL 189 190 A combined date and time at which the layer was created, formatted as defined by [RFC 3339, section 5.6][rfc3339-s5.6]. 191 192 - **author** *string*, OPTIONAL 193 194 The author of the build point. 195 196 - **created_by** *string*, OPTIONAL 197 198 The command which created the layer. 199 200 - **comment** *string*, OPTIONAL 201 202 A custom message set when creating the layer. 203 204 - **empty_layer** *boolean*, OPTIONAL 205 206 This field is used to mark if the history item created a filesystem diff. 207 It is set to true if this history item doesn't correspond to an actual layer in the rootfs section (for example, Dockerfile's [ENV](https://docs.docker.com/engine/reference/builder/#/env) command results in no change to the filesystem). 208 209Any extra fields in the Image JSON struct are considered implementation specific and MUST be ignored by any implementations which are unable to interpret them. 210 211Whitespace is OPTIONAL and implementations MAY have compact JSON with no whitespace. 212 213## Example 214 215Here is an example image configuration JSON document: 216 217```json,title=Image%20JSON&mediatype=application/vnd.oci.image.config.v1%2Bjson 218{ 219 "created": "2015-10-31T22:22:56.015925234Z", 220 "author": "Alyssa P. Hacker <alyspdev@example.com>", 221 "architecture": "amd64", 222 "os": "linux", 223 "config": { 224 "User": "alice", 225 "ExposedPorts": { 226 "8080/tcp": {} 227 }, 228 "Env": [ 229 "PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin", 230 "FOO=oci_is_a", 231 "BAR=well_written_spec" 232 ], 233 "Entrypoint": [ 234 "/bin/my-app-binary" 235 ], 236 "Cmd": [ 237 "--foreground", 238 "--config", 239 "/etc/my-app.d/default.cfg" 240 ], 241 "Volumes": { 242 "/var/job-result-data": {}, 243 "/var/log/my-app-logs": {} 244 }, 245 "WorkingDir": "/home/alice", 246 "Labels": { 247 "com.example.project.git.url": "https://example.com/project.git", 248 "com.example.project.git.commit": "45a939b2999782a3f005621a8d0f29aa387e1d6b" 249 } 250 }, 251 "rootfs": { 252 "diff_ids": [ 253 "sha256:c6f988f4874bb0add23a778f753c65efe992244e148a1d2ec2a8b664fb66bbd1", 254 "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef" 255 ], 256 "type": "layers" 257 }, 258 "history": [ 259 { 260 "created": "2015-10-31T22:22:54.690851953Z", 261 "created_by": "/bin/sh -c #(nop) ADD file:a3bc1e842b69636f9df5256c49c5374fb4eef1e281fe3f282c65fb853ee171c5 in /" 262 }, 263 { 264 "created": "2015-10-31T22:22:55.613815829Z", 265 "created_by": "/bin/sh -c #(nop) CMD [\"sh\"]", 266 "empty_layer": true 267 } 268 ] 269} 270``` 271 272[rfc3339-s5.6]: https://tools.ietf.org/html/rfc3339#section-5.6 273[go-environment]: https://golang.org/doc/install/source#environment 274[tar-split]: https://github.com/vbatts/tar-split 275