api: add method to load BlobCacheConfigV2 from file

jiangliu · jiangliu · commit f9b051ed400b · 2023-03-05T01:39:25.000+08:00
Add method to load BlobCacheConfigV2 from configuration file.

Signed-off-by: Jiang Liu &lt;gerry@linux.alibaba.com&gt;
diff --git a/api/src/config.rs b/api/src/config.rs
@@ -890,11 +890,54 @@ pub struct BlobCacheEntryConfigV2 {
     /// Configuration information for local cache system.
     #[serde(default)]
     pub cache: CacheConfigV2,
-    /// Optional file path for metadata blobs.
+    /// Optional file path for metadata blob.
     #[serde(default)]
     pub metadata_path: Option<String>,
 }
 
+impl BlobCacheEntryConfigV2 {
+    /// Read configuration information from a file.
+    pub fn from_file<P: AsRef<Path>>(path: P) -> Result<Self> {
+        let md = fs::metadata(path.as_ref())?;
+        if md.len() > 0x100000 {
+            return Err(eother!("configuration file size is too big"));
+        }
+        let content = fs::read_to_string(path)?;
+        Self::from_str(&content)
+    }
+
+    /// Validate the configuration object.
+    pub fn validate(&self) -> bool {
+        if self.version != 2 {
+            return false;
+        }
+        let config: ConfigV2 = self.into();
+        config.validate()
+    }
+}
+
+impl FromStr for BlobCacheEntryConfigV2 {
+    type Err = Error;
+
+    fn from_str(s: &str) -> Result<BlobCacheEntryConfigV2> {
+        if let Ok(v) = serde_json::from_str::<BlobCacheEntryConfigV2>(s) {
+            return if v.validate() {
+                Ok(v)
+            } else {
+                Err(einval!("invalid configuration"))
+            };
+        }
+        if let Ok(v) = toml::from_str::<BlobCacheEntryConfigV2>(s) {
+            return if v.validate() {
+                Ok(v)
+            } else {
+                Err(einval!("invalid configuration"))
+            };
+        }
+        Err(einval!("failed to parse configuration information"))
+    }
+}
+
 impl From<&BlobCacheEntryConfigV2> for ConfigV2 {
     fn from(c: &BlobCacheEntryConfigV2) -> Self {
         ConfigV2 {
@@ -943,6 +986,106 @@ impl ConfigV2Internal {
     }
 }
 
+/// Blob cache object type for nydus/rafs bootstrap blob.
+pub const BLOB_CACHE_TYPE_META_BLOB: &str = "bootstrap";
+/// Blob cache object type for nydus/rafs data blob.
+pub const BLOB_CACHE_TYPE_DATA_BLOB: &str = "datablob";
+
+/// Configuration information for a cached blob.
+#[derive(Debug, Deserialize, Serialize)]
+pub struct BlobCacheEntry {
+    /// Type of blob object, bootstrap or data blob.
+    #[serde(rename = "type")]
+    pub blob_type: String,
+    /// Blob id.
+    #[serde(rename = "id")]
+    pub blob_id: String,
+    /// Configuration information to generate blob cache object.
+    #[serde(default, rename = "config")]
+    pub(crate) blob_config_legacy: Option<BlobCacheEntryConfig>,
+    /// Configuration information to generate blob cache object.
+    #[serde(default, rename = "config_v2")]
+    pub blob_config: Option<BlobCacheEntryConfigV2>,
+    /// Domain id for the blob, which is used to group cached blobs into management domains.
+    #[serde(default)]
+    pub domain_id: String,
+}
+
+impl BlobCacheEntry {
+    pub fn prepare_configuration_info(&mut self) -> bool {
+        if self.blob_config.is_none() {
+            if let Some(legacy) = self.blob_config_legacy.as_ref() {
+                match legacy.try_into() {
+                    Err(_) => return false,
+                    Ok(v) => self.blob_config = Some(v),
+                }
+            }
+        }
+
+        match self.blob_config.as_ref() {
+            None => false,
+            Some(cfg) => cfg.cache.validate() && cfg.backend.validate(),
+        }
+    }
+}
+
+impl BlobCacheEntry {
+    /// Read configuration information from a file.
+    pub fn from_file<P: AsRef<Path>>(path: P) -> Result<Self> {
+        let md = fs::metadata(path.as_ref())?;
+        if md.len() > 0x100000 {
+            return Err(eother!("configuration file size is too big"));
+        }
+        let content = fs::read_to_string(path)?;
+        Self::from_str(&content)
+    }
+
+    /// Validate the configuration object.
+    pub fn validate(&self) -> bool {
+        if self.blob_type != BLOB_CACHE_TYPE_META_BLOB
+            && self.blob_type != BLOB_CACHE_TYPE_DATA_BLOB
+        {
+            warn!("invalid blob type {} for blob cache entry", self.blob_type);
+            return false;
+        }
+        if let Some(config) = self.blob_config.as_ref() {
+            if !config.validate() {
+                return false;
+            }
+        }
+        true
+    }
+}
+
+impl FromStr for BlobCacheEntry {
+    type Err = Error;
+
+    fn from_str(s: &str) -> Result<BlobCacheEntry> {
+        if let Ok(v) = serde_json::from_str::<BlobCacheEntry>(s) {
+            return if v.validate() {
+                Ok(v)
+            } else {
+                Err(einval!("invalid configuration"))
+            };
+        }
+        if let Ok(v) = toml::from_str::<BlobCacheEntry>(s) {
+            return if v.validate() {
+                Ok(v)
+            } else {
+                Err(einval!("invalid configuration"))
+            };
+        }
+        Err(einval!("failed to parse configuration information"))
+    }
+}
+
+/// Configuration information for a list of cached blob objects.
+#[derive(Debug, Default, Deserialize, Serialize)]
+pub struct BlobCacheList {
+    /// List of blob configuration information.
+    pub blobs: Vec<BlobCacheEntry>,
+}
+
 fn default_true() -> bool {
     true
 }
diff --git a/api/src/http.rs b/api/src/http.rs
@@ -4,15 +4,14 @@
 //
 // SPDX-License-Identifier: Apache-2.0
 
-use std::convert::TryInto;
 use std::io;
 use std::sync::mpsc::{RecvError, SendError};
 
 use nydus_error::error::MetricsError;
 use serde::Deserialize;
 use serde_json::Error as SerdeError;
 
-use crate::{BlobCacheEntryConfig, BlobCacheEntryConfigV2};
+use crate::BlobCacheEntry;
 
 /// Mount a filesystem.
 #[derive(Clone, Deserialize, Debug)]
@@ -43,56 +42,6 @@ pub struct DaemonConf {
     pub log_level: String,
 }
 
-/// Blob cache object type for nydus/rafs bootstrap blob.
-pub const BLOB_CACHE_TYPE_META_BLOB: &str = "bootstrap";
-/// Blob cache object type for nydus/rafs data blob.
-pub const BLOB_CACHE_TYPE_DATA_BLOB: &str = "datablob";
-
-/// Configuration information for a cached blob.
-#[derive(Debug, Deserialize, Serialize)]
-pub struct BlobCacheEntry {
-    /// Type of blob object, bootstrap or data blob.
-    #[serde(rename = "type")]
-    pub blob_type: String,
-    /// Blob id.
-    #[serde(rename = "id")]
-    pub blob_id: String,
-    /// Configuration information to generate blob cache object.
-    #[serde(default, rename = "config")]
-    pub(crate) blob_config_legacy: Option<BlobCacheEntryConfig>,
-    /// Configuration information to generate blob cache object.
-    #[serde(default, rename = "config_v2")]
-    pub blob_config: Option<BlobCacheEntryConfigV2>,
-    /// Domain id for the blob, which is used to group cached blobs into management domains.
-    #[serde(default)]
-    pub domain_id: String,
-}
-
-impl BlobCacheEntry {
-    pub fn prepare_configuration_info(&mut self) -> bool {
-        if self.blob_config.is_none() {
-            if let Some(legacy) = self.blob_config_legacy.as_ref() {
-                match legacy.try_into() {
-                    Err(_) => return false,
-                    Ok(v) => self.blob_config = Some(v),
-                }
-            }
-        }
-
-        match self.blob_config.as_ref() {
-            None => false,
-            Some(cfg) => cfg.cache.validate() && cfg.backend.validate(),
-        }
-    }
-}
-
-/// Configuration information for a list of cached blob objects.
-#[derive(Debug, Default, Deserialize, Serialize)]
-pub struct BlobCacheList {
-    /// List of blob configuration information.
-    pub blobs: Vec<BlobCacheEntry>,
-}
-
 /// Identifier for cached blob objects.
 ///
 /// Domains are used to control the blob sharing scope. All blobs associated with the same domain
diff --git a/docs/samples/blob_cache_entry.toml b/docs/samples/blob_cache_entry.toml
@@ -0,0 +1,145 @@
+# Configuration file for Nydus Image Service
+
+type = "bootstrap"
+id = "image1"
+domain_id = "domain1"
+
+# Configuration file format version number, must be 2.
+[config_v2]
+version = 2
+# Identifier for the instance.
+id = "my_id"
+# Optional file path for metadata blobs, for BlobCacheEntry only.
+metadata_path = "/path/to/rafs/meta/data/blob"
+
+[config_v2.backend]
+# Type of storage backend, valid values: "localfs", "oss", "registry"
+type = "localfs"
+
+[config_v2.backend.localfs]
+blob_file = "/tmp/nydus.blob.data"
+dir = "/tmp"
+alt_dirs = ["/var/nydus/cache"]
+
+[config_v2.backend.oss]
+# Oss http scheme, either 'http' or 'https'
+scheme = "http"
+# Oss endpoint
+endpoint = "my_endpoint"
+# Oss bucket name
+bucket_name = "my_bucket_name"
+# Prefix object_prefix to OSS object key, for example the simulation of subdirectory:
+object_prefix = "my_object_prefix"
+# Oss access key
+access_key_id = "my_access_key_id"
+# Oss secret
+access_key_secret = "my_access_key_secret"
+# Skip SSL certificate validation for HTTPS scheme.
+skip_verify = true
+# Drop the read request once http request timeout, in seconds.
+timeout = 10
+# Drop the read request once http connection timeout, in seconds.
+connect_timeout = 10
+# Retry count when read request failed.
+retry_limit = 5
+
+[config_v2.backend.oss.proxy]
+# Access remote storage backend via proxy, e.g. Dragonfly dfdaemon server URL.
+url = "localhost:6789"
+# Proxy health checking endpoint.
+ping_url = "localhost:6789/ping"
+# Fallback to remote storage backend if proxy ping failed.
+fallback = true
+# Interval for proxy health checking, in seconds.
+check_interval = 5
+# Replace URL to http to request source registry with proxy, and allow fallback to https if the proxy is unhealthy.
+use_http = false
+
+[[config_v2.backend.oss.mirrors]]
+# Mirror server URL, for example http://127.0.0.1:65001.
+host = "http://127.0.0.1:65001"
+# Ping URL to check mirror server health.
+ping_url = "http://127.0.0.1:65001/ping"
+# HTTP request headers to be passed to mirror server.
+# headers =
+# Whether the authorization process is through mirror, default to false.
+auth_through = true
+# Interval for mirror health checking, in seconds.
+health_check_interval = 5
+# Maximum number of failures before marking a mirror as unusable.
+failure_limit = 5
+
+[config_v2.backend.registry]
+# Registry http scheme, either 'http' or 'https'
+scheme = "https"
+# Registry url host
+host = "my.registry.com"
+# Registry image name, like 'library/ubuntu'
+repo = "nydus"
+# Base64_encoded(username:password), the field should be sent to registry auth server to get a bearer token.
+auth = "base64_encoded"
+# Skip SSL certificate validation for HTTPS scheme.
+skip_verify = true
+# Drop the read request once http request timeout, in seconds.
+timeout = 10
+# Drop the read request once http connection timeout, in seconds.
+connect_timeout = 10
+# Retry count when read request failed.
+retry_limit = 5
+# The field is a bearer token to be sent to registry to authorize registry requests.
+registry_token = "bear_token"
+# The http scheme to access blobs.
+# It is used to workaround some P2P subsystem that requires a different scheme than the registry.
+blob_url_scheme = "https"
+# Redirect blob access to a different host regardless of the one specified in 'host'.
+blob_redirected_host = "redirect.registry.com"
+
+[config_v2.backend.registry.proxy]
+# Access remote storage backend via proxy, e.g. Dragonfly dfdaemon server URL.
+url = "localhost:6789"
+# Proxy health checking endpoint.
+ping_url = "localhost:6789/ping"
+# Fallback to remote storage backend if proxy ping failed.
+fallback = true
+# Interval for proxy health checking, in seconds.
+check_interval = 5
+# Replace URL to http to request source registry with proxy, and allow fallback to https if the proxy is unhealthy.
+use_http = false
+
+[[config_v2.backend.registry.mirrors]]
+# Mirror server URL, for example http://127.0.0.1:65001.
+host = "http://127.0.0.1:65001"
+# Ping URL to check mirror server health.
+ping_url = "http://127.0.0.1:65001/ping"
+# HTTP request headers to be passed to mirror server.
+# headers =
+# Whether the authorization process is through mirror, default to false.
+auth_through = true
+# Interval for mirror health checking, in seconds.
+health_check_interval = 5
+# Maximum number of failures before marking a mirror as unusable.
+failure_limit = 5
+
+[config_v2.cache]
+# Type of blob cache: "blobcache", "filecache", "fscache", "dummycache" or ""
+type = "filecache"
+# Whether to cache compressed or uncompressed data.
+compressed = true
+# Whether to validate data read from the cache.
+validate = true
+
+[config_v2.cache.filecache]
+work_dir = "."
+
+[config_v2.cache.fscache]
+work_dir = "."
+
+[config_v2.cache.prefetch]
+# Whether to enable blob data prefetching.
+enable = true
+# Number of data prefetching working threads, valid values: 1-1024.
+threads = 8
+# The batch size to prefetch data from backend, valid values: 0-0x10000000.
+batch_size = 1000000
+# Network bandwidth rate limit in unit of Bytes and Zero means no limit.
+bandwidth_limit = 10000000
diff --git a/service/src/blob_cache.rs b/service/src/blob_cache.rs
@@ -328,9 +328,12 @@ impl BlobCacheMgr {
                 "blob_cache: `config.metadata_path` for meta blob is empty"
             ));
         }
-        let path = Path::new(&path)
-            .canonicalize()
-            .map_err(|_e| einval!("blob_cache: `config.metadata_path` for meta blob is invalid"))?;
+        let path = Path::new(&path).canonicalize().map_err(|_e| {
+            einval!(format!(
+                "blob_cache: `config.metadata_path={}` for meta blob is invalid",
+                path
+            ))
+        })?;
         if !path.is_file() {
             return Err(einval!(
                 "blob_cache: `config.metadata_path` for meta blob is not a file"
diff --git a/service/src/singleton.rs b/service/src/singleton.rs
diff --git a/src/bin/nydusd/main.rs b/src/bin/nydusd/main.rs
