add instance_id generated at install time (#245)

docs/telemetry.md

@@ -66,6 +66,7 @@ The following describes the information sent to Microsoft if telemetry is enable
 
 The following are common data types used in multiple locations:
 
+* Instance ID - A randomly generated GUID used to uniquely identify an instance of OneFuzz
 * Task ID - A randomly generated GUID used to uniquely identify a fuzzing task.
 * Job ID - A randomly generated GUID used to uniquely identify a job.
 * Machine ID - A GUID used to identify the machine running the task. When run in

src/agent/onefuzz-agent/src/debug/generic_crash_report.rs

@@ -64,6 +64,7 @@ pub fn run(args: &clap::ArgMatches) -> Result<()> {
             telemetry_key: None,
             job_id: Uuid::parse_str("00000000-0000-0000-0000-000000000000").unwrap(),
             task_id: Uuid::parse_str("11111111-1111-1111-1111-111111111111").unwrap(),
+            instance_id: Uuid::parse_str("22222222-2222-2222-2222-222222222222").unwrap(),
         },
     };
 

src/agent/onefuzz-agent/src/debug/libfuzzer_coverage.rs

@@ -64,6 +64,7 @@ pub fn run(args: &clap::ArgMatches) -> Result<()> {
             telemetry_key: None,
             job_id: Uuid::parse_str("00000000-0000-0000-0000-000000000000").unwrap(),
             task_id: Uuid::parse_str("11111111-1111-1111-1111-111111111111").unwrap(),
+            instance_id: Uuid::parse_str("22222222-2222-2222-2222-222222222222").unwrap(),
         },
     };
 

src/agent/onefuzz-agent/src/debug/libfuzzer_crash_report.rs

@@ -60,6 +60,7 @@ pub fn run(args: &clap::ArgMatches) -> Result<()> {
             telemetry_key: None,
             job_id: Uuid::parse_str("00000000-0000-0000-0000-000000000000").unwrap(),
             task_id: Uuid::parse_str("11111111-1111-1111-1111-111111111111").unwrap(),
+            instance_id: Uuid::parse_str("22222222-2222-2222-2222-222222222222").unwrap(),
         },
     };
 

src/agent/onefuzz-agent/src/debug/libfuzzer_fuzz.rs

@@ -62,6 +62,7 @@ pub fn run(args: &clap::ArgMatches) -> Result<()> {
             telemetry_key: None,
             job_id: Uuid::parse_str("00000000-0000-0000-0000-000000000000").unwrap(),
             task_id: Uuid::parse_str("11111111-1111-1111-1111-111111111111").unwrap(),
+            instance_id: Uuid::parse_str("22222222-2222-2222-2222-222222222222").unwrap(),
         },
     };
 

src/agent/onefuzz-agent/src/tasks/config.rs

@@ -26,6 +26,8 @@ pub struct CommonConfig {
 
     pub task_id: Uuid,
 
+    pub instance_id: Uuid,
+
     pub instrumentation_key: Option<Uuid>,
 
     pub heartbeat_queue: Option<Url>,
@@ -127,6 +129,7 @@ impl Config {
         telemetry::set_property(EventData::TaskId(self.common().task_id));
         telemetry::set_property(EventData::MachineId(get_machine_id().await?));
         telemetry::set_property(EventData::Version(env!("ONEFUZZ_VERSION").to_string()));
+        telemetry::set_property(EventData::InstanceId(self.common().instance_id));
         if let Ok(scaleset) = get_scaleset_name().await {
             telemetry::set_property(EventData::ScalesetId(scaleset));
         }

src/agent/onefuzz-supervisor/src/config.rs

@@ -28,6 +28,8 @@ pub struct StaticConfig {
     pub telemetry_key: Option<Uuid>,
 
     pub heartbeat_queue: Option<Url>,
+
+    pub instance_id: Uuid,
 }
 
 // Temporary shim type to bridge the current service-provided config.
@@ -44,6 +46,8 @@ struct RawStaticConfig {
     pub telemetry_key: Option<Uuid>,
 
     pub heartbeat_queue: Option<Url>,
+
+    pub instance_id: Uuid,
 }
 
 impl StaticConfig {
@@ -70,6 +74,7 @@ impl StaticConfig {
             instrumentation_key: config.instrumentation_key,
             telemetry_key: config.telemetry_key,
             heartbeat_queue: config.heartbeat_queue,
+            instance_id: config.instance_id,
         };
 
         Ok(config)

src/agent/onefuzz-supervisor/src/main.rs

@@ -120,6 +120,7 @@ fn load_config(opt: RunOpt) -> Result<StaticConfig> {
 }
 
 async fn run_agent(config: StaticConfig) -> Result<()> {
+    telemetry::set_property(EventData::InstanceId(config.instance_id));
     telemetry::set_property(EventData::MachineId(get_machine_id().await?));
     telemetry::set_property(EventData::Version(env!("ONEFUZZ_VERSION").to_string()));
     if let Ok(scaleset) = get_scaleset_name().await {

src/agent/onefuzz/src/telemetry.rs

@@ -43,6 +43,7 @@ impl Event {
 #[derive(Clone, Debug, PartialEq)]
 pub enum EventData {
     WorkerId(u64),
+    InstanceId(Uuid),
     JobId(Uuid),
     TaskId(Uuid),
     ScalesetId(String),
@@ -77,6 +78,7 @@ impl EventData {
     pub fn as_values(&self) -> (&str, String) {
         match self {
             Self::Version(x) => ("version", x.to_string()),
+            Self::InstanceId(x) => ("instance_id", x.to_string()),
             Self::JobId(x) => ("job_id", x.to_string()),
             Self::TaskId(x) => ("task_id", x.to_string()),
             Self::ScalesetId(x) => ("scaleset_id", x.to_string()),
@@ -113,6 +115,7 @@ impl EventData {
             // TODO: Request CELA review of Version, as having this for central stats
             //       would be useful to track uptake of new releases
             Self::Version(_) => false,
+            Self::InstanceId(_) => true,
             Self::TaskId(_) => true,
             Self::JobId(_) => true,
             Self::MachineId(_) => true,

src/api-service/__app__/info/__init__.py

@@ -9,6 +9,7 @@ from onefuzztypes.responses import Info
 from ..onefuzzlib.azure.creds import (
     get_base_region,
     get_base_resource_group,
+    get_instance_id,
     get_subscription,
 )
 from ..onefuzzlib.request import ok
@@ -22,5 +23,6 @@ def main(req: func.HttpRequest) -> func.HttpResponse:
             region=get_base_region(),
             subscription=get_subscription(),
             versions=versions(),
+            instance_id=get_instance_id(),
         )
     )

src/api-service/__app__/onefuzzlib/azure/containers.py

@@ -6,7 +6,7 @@
 import datetime
 import os
 import urllib.parse
-from typing import Any, Dict, Optional, Union, cast
+from typing import Dict, Optional, Union, cast
 
 from azure.common import AzureHttpError, AzureMissingResourceHttpError
 from azure.storage.blob import BlobPermissions, ContainerPermissions
@@ -129,11 +129,11 @@ def save_blob(
 
 def get_blob(
     container: str, name: str, account_id: Optional[str] = None
-) -> Optional[Any]:  # should be bytes
+) -> Optional[bytes]:
     service = get_blob_service(account_id)
     try:
         blob = service.get_blob_to_bytes(container, name).content
-        return blob
+        return cast(bytes, blob)
     except AzureMissingResourceHttpError:
         return None
 

src/api-service/__app__/onefuzzlib/azure/creds.py

@@ -102,6 +102,16 @@ def get_instance_url() -> str:
     return "https://%s.azurewebsites.net" % get_instance_name()
 
 
+@cached
+def get_instance_id() -> UUID:
+    from .containers import get_blob
+
+    blob = get_blob("base-config", "instance_id", account_id=get_func_storage())
+    if blob is None:
+        raise Exception("missing instance_id")
+    return UUID(blob.decode())
+
+
 DAY_IN_SECONDS = 60 * 60 * 24
 
 

src/api-service/__app__/onefuzzlib/extension.py

@@ -12,7 +12,7 @@ from onefuzztypes.models import AgentConfig, ReproConfig
 from onefuzztypes.primitives import Extension, Region
 
 from .azure.containers import get_container_sas_url, get_file_sas_url, save_blob
-from .azure.creds import get_func_storage, get_instance_url
+from .azure.creds import get_func_storage, get_instance_id, get_instance_url
 from .azure.monitor import get_monitor_settings
 from .azure.queue import get_queue_sas
 from .reports import get_report
@@ -100,6 +100,7 @@ def build_pool_config(pool_name: str) -> str:
             add=True,
         ),
         telemetry_key=os.environ.get("ONEFUZZ_TELEMETRY"),
+        instance_id=get_instance_id(),
     )
 
     save_blob(

src/api-service/__app__/onefuzzlib/reports.py

@@ -5,7 +5,7 @@
 
 import json
 import logging
-from typing import Optional
+from typing import Optional, Union
 
 from onefuzztypes.models import Report
 from pydantic import ValidationError
@@ -13,7 +13,9 @@ from pydantic import ValidationError
 from .azure.containers import get_blob
 
 
-def parse_report(content: str, metadata: Optional[str] = None) -> Optional[Report]:
+def parse_report(
+    content: Union[str, bytes], metadata: Optional[str] = None
+) -> Optional[Report]:
     if isinstance(content, bytes):
         try:
             content = content.decode()

src/api-service/__app__/onefuzzlib/tasks/config.py

@@ -12,7 +12,12 @@ from onefuzztypes.enums import Compare, ContainerPermission, ContainerType, Task
 from onefuzztypes.models import TaskConfig, TaskDefinition, TaskUnitConfig
 
 from ..azure.containers import blob_exists, container_exists, get_container_sas_url
-from ..azure.creds import get_func_storage, get_fuzz_storage, get_instance_url
+from ..azure.creds import (
+    get_func_storage,
+    get_fuzz_storage,
+    get_instance_id,
+    get_instance_url,
+)
 from ..azure.queue import get_queue_sas
 from .defs import TASK_DEFINITIONS
 
@@ -187,6 +192,7 @@ def build_task_config(
             add=True,
         ),
         back_channel_address="https://%s/api/back_channel" % (get_instance_url()),
+        instance_id=get_instance_id(),
     )
 
     if definition.monitor_queue:

src/api-service/__app__/pool/__init__.py

@@ -15,6 +15,7 @@ from onefuzztypes.responses import BoolResult
 from ..onefuzzlib.azure.creds import (
     get_base_region,
     get_func_storage,
+    get_instance_id,
     get_instance_url,
     get_regions,
 )
@@ -35,6 +36,7 @@ def set_config(pool: Pool) -> Pool:
             account_id=get_func_storage(),
             add=True,
         ),
+        instance_id=get_instance_id(),
     )
     return pool
 

src/deployment/azuredeploy.json

@@ -63,8 +63,7 @@
             "namespace": "onefuzz",
             "members": {
                 "severitiesAtMostInfo": {
-                    "parameters": [
+                    "parameters": [],
-                    ],
                     "output": {
                         "type": "array",
                         "value": [
@@ -209,8 +208,7 @@
                     ],
                     "linuxFxVersion": "Python|3.7",
                     "alwaysOn": true,
-                    "defaultDocuments": [
+                    "defaultDocuments": [],
-                    ],
                     "httpLoggingEnabled": true,
                     "logsDirectorySizeLimit": 100,
                     "detailedErrorLoggingEnabled": true,
@@ -243,8 +241,7 @@
             "type": "Microsoft.Web/serverFarms",
             "location": "[resourceGroup().location]",
             "kind": "linux",
-            "dependsOn": [
+            "dependsOn": [],
-            ],
             "properties": {
                 "name": "[parameters('name')]",
                 "reserved": true
@@ -394,7 +391,6 @@
                     "properties": {
                         "workspaceResourceId": "[resourceId('Microsoft.OperationalInsights/workspaces', variables('monitorAccountName'))]"
                     },
-
                     "plan": {
                         "name": "[concat('VMInsights', '(', variables('monitorAccountName'), ')')]",
                         "publisher": "Microsoft",
@@ -694,20 +690,17 @@
                     {
                         "flag": "ServiceMode",
                         "value": "Serverless",
-                        "properties": {
+                        "properties": {}
-                        }
                     },
                     {
                         "flag": "EnableConnectivityLogs",
                         "value": "True",
-                        "properties": {
+                        "properties": {}
-                        }
                     },
                     {
                         "flag": "EnableMessagingLogs",
                         "value": "False",
-                        "properties": {
+                        "properties": {}
-                        }
                     }
                 ]
             }
@@ -743,4 +736,4 @@
             "value": "[variables('scaleset_identity')]"
         }
     }
-}
+}

src/deployment/deploy.py

@@ -92,6 +92,7 @@ FUNC_TOOLS_ERROR = (
 
 logger = logging.getLogger("deploy")
 
+
 def gen_guid():
     return str(uuid.uuid4())
 
@@ -464,6 +465,29 @@ class Client:
                 % json.dumps(result.as_dict(), indent=4, sort_keys=True),
             )
 
+    def add_instance_id(self):
+        logger.info("setting instance_id log export")
+
+        container_name = "base-config"
+        blob_name = "instance_id"
+        account_name = self.results["deploy"]["func-name"]["value"]
+        key = self.results["deploy"]["func-key"]["value"]
+        account_url = "https://%s.blob.core.windows.net" % account_name
+        client = BlobServiceClient(account_url, credential=key)
+        if container_name not in [x["name"] for x in client.list_containers()]:
+            client.create_container(container_name)
+
+        blob_client = client.get_blob_client(container_name, blob_name)
+        if blob_client.exists():
+            logger.debug("instance_id already exists")
+            instance_id = uuid.UUID(blob_client.download_blob().readall())
+        else:
+            logger.debug("creating new instance_id")
+            instance_id = uuid.uuid4()
+            blob_client.upload_blob(str(instance_id))
+
+        logger.info("instance_id: %s", instance_id)
+
     def add_log_export(self):
         if not self.export_appinsights:
             logger.info("not exporting appinsights")
@@ -683,6 +707,7 @@ def main():
         ("queues", Client.create_queues),
         ("eventgrid", Client.create_eventgrid),
         ("tools", Client.upload_tools),
+        ("add_instance_id", Client.add_instance_id),
         ("instance-specific-setup", Client.upload_instance_setup),
         ("third-party", Client.upload_third_party),
         ("api", Client.deploy_app),

src/pytypes/onefuzztypes/models.py

@@ -270,9 +270,11 @@ class AgentConfig(BaseModel):
     heartbeat_queue: Optional[str]
     instrumentation_key: Optional[str]
     telemetry_key: Optional[str]
+    instance_id: UUID
 
 
 class TaskUnitConfig(BaseModel):
+    instance_id: UUID
     job_id: UUID
     task_id: UUID
     task_type: TaskType

src/pytypes/onefuzztypes/responses.py

@@ -4,6 +4,7 @@
 # Licensed under the MIT License.
 
 from typing import Dict, Optional
+from uuid import UUID
 
 from pydantic import BaseModel
 
@@ -35,6 +36,7 @@ class Info(BaseResponse):
     region: Region
     subscription: str
     versions: Dict[str, Version]
+    instance_id: Optional[UUID]
 
 
 class ContainerInfoBase(BaseResponse):