log analytics (#1851)

Co-authored-by: stas <statis@microsoft.com>

src/ApiService/ApiService/ApiService.csproj

@@ -8,12 +8,14 @@
     <WarningLevel>5</WarningLevel>
   </PropertyGroup>
   <ItemGroup>
+    <PackageReference Include="Azure.ResourceManager.Monitor" Version="1.0.0-beta.2" />
     <PackageReference Include="Faithlife.Utility" Version="0.12.2" />
     <PackageReference Include="Azure.Security.KeyVault.Secrets" Version="4.3.0" />
     <PackageReference Include="Microsoft.Azure.Functions.Worker.Extensions.Storage" Version="5.0.0" />
     <PackageReference Include="Microsoft.Azure.Functions.Worker.Extensions.EventGrid" Version="2.1.0" />
     <PackageReference Include="Microsoft.Azure.Functions.Worker.Extensions.Http" Version="3.0.13" />
     <PackageReference Include="Microsoft.Azure.Functions.Worker.Extensions.Timer" Version="4.1.0" />
+    <PackageReference Include="Microsoft.Azure.Management.OperationalInsights" Version="0.24.0-preview" />
     <PackageReference Include="Microsoft.Extensions.Logging.ApplicationInsights" Version="2.20.0" />
     <PackageReference Include="Microsoft.Azure.Functions.Worker.Sdk" Version="1.3.0" OutputItemType="Analyzer" />
     <PackageReference Include="Microsoft.Azure.Functions.Worker" Version="1.6.0" />

src/ApiService/ApiService/OneFuzzTypes/Enums.cs

@@ -295,3 +295,10 @@ public static PoolState[] Available() {
 public enum Architecture {
     x86_64
 }
+
+
+public enum AgentMode {
+    Fuzz,
+    Repro,
+    Proxy
+}

src/ApiService/ApiService/OneFuzzTypes/Events.cs

@@ -57,6 +57,9 @@ public EventType GetEventType() {
                 EventTaskFailed _ => EventType.TaskFailed,
                 EventTaskStopped _ => EventType.TaskStopped,
                 EventTaskStateUpdated _ => EventType.TaskStateUpdated,
+                EventScalesetFailed _ => EventType.ScalesetFailed,
+                EventScalesetResizeScheduled _ => EventType.ScalesetResizeScheduled,
+                EventScalesetStateUpdated _ => EventType.ScalesetStateUpdated,
                 _ => throw new NotImplementedException(),
             };
 
@@ -166,11 +169,11 @@ Guid PingId
 //    int Size) : BaseEvent();
 
 
-//record EventScalesetFailed(
-//    Guid ScalesetId,
-//    PoolName: PoolName,
-//    Error: Error
-//): BaseEvent();
+public record EventScalesetFailed(
+    Guid ScalesetId,
+    PoolName PoolName,
+    Error Error
+) : BaseEvent();
 
 
 //record EventScalesetDeleted(
@@ -180,11 +183,11 @@ Guid PingId
 //    ) : BaseEvent();
 
 
-//record EventScalesetResizeScheduled(
-//    Guid ScalesetId,
-//    PoolName PoolName,
-//    int size
-//    ) : BaseEvent();
+public record EventScalesetResizeScheduled(
+    Guid ScalesetId,
+    PoolName PoolName,
+    int size
+    ) : BaseEvent();
 
 
 //record EventPoolDeleted(
@@ -249,11 +252,11 @@ PoolName PoolName
 //        ) : BaseEvent();
 
 
-//    record EventScalesetStateUpdated(
-//        Guid ScalesetId,
-//        PoolName PoolName,
-//        ScalesetState State
-//        ) : BaseEvent();
+public record EventScalesetStateUpdated(
+    Guid ScalesetId,
+    PoolName PoolName,
+    ScalesetState State
+) : BaseEvent();
 
 //    record EventNodeStateUpdated(
 //        Guid MachineId,

src/ApiService/ApiService/OneFuzzTypes/Model.cs

@@ -386,6 +386,7 @@ public record Scaleset(
     bool SpotInstance,
     bool EphemeralOsDisks,
     bool NeedsConfigUpdate,
+    Error? Error,
     List<ScalesetNodeState> Nodes,
     Guid? ClientId,
     Guid? ClientObjectId,
@@ -513,8 +514,26 @@ public record Pool(
 ) : StatefulEntityBase<PoolState>(State);
 
 
-// TODO
-public record AgentConfig();
+public record ClientCredentials
+(
+    Guid ClientId,
+    string ClientSecret
+);
+
+
+public record AgentConfig(
+    ClientCredentials? ClientCredentials,
+    [property: JsonPropertyName("onefuzz_url")] Uri OneFuzzUrl,
+    PoolName PoolName,
+    Uri? HeartbeatQueue,
+    string? InstanceTelemetryKey,
+    string? MicrosoftTelemetryKey,
+    string? MultiTenantDomain,
+    Guid InstanceId
+);
+
+
+
 public record WorkSetSummary();
 public record ScalesetSummary();
 

src/ApiService/ApiService/Program.cs

@@ -1,8 +1,11 @@
 // to avoid collision with Task in model.cs
 global using System;
-global using System.Collections.Generic;
-global using System.Linq;
-global using Async = System.Threading.Tasks;
+global
+using System.Collections.Generic;
+global
+using System.Linq;
+global
+using Async = System.Threading.Tasks;
 using Microsoft.Azure.Functions.Worker;
 using Microsoft.Azure.Functions.Worker.Middleware;
 using Microsoft.Extensions.DependencyInjection;
@@ -81,6 +84,7 @@ public static void Main() {
             .AddScoped<IJobOperations, JobOperations>()
             .AddScoped<IScheduler, Scheduler>()
             .AddScoped<IConfig, Config>()
+            .AddScoped<ILogAnalytics, LogAnalytics>()
 
             //Move out expensive resources into separate class, and add those as Singleton
             // ArmClient, Table Client(s), Queue Client(s), HttpClient, etc.\

src/ApiService/ApiService/onefuzzlib/Containers.cs

@@ -1,35 +1,50 @@
-using System.Threading.Tasks;
-using Azure;
+using Azure;
 using Azure.ResourceManager;
 using Azure.Storage;
 using Azure.Storage.Blobs;
 using Azure.Storage.Sas;
 
+
 namespace Microsoft.OneFuzz.Service;
 
 
 public interface IContainers {
-    public Task<BinaryData?> GetBlob(Container container, string name, StorageType storageType);
+    public Async.Task<BinaryData?> GetBlob(Container container, string name, StorageType storageType);
 
     public Async.Task<BlobContainerClient?> FindContainer(Container container, StorageType storageType);
 
     public Async.Task<Uri?> GetFileSasUrl(Container container, string name, StorageType storageType, BlobSasPermissions permissions, TimeSpan? duration = null);
-    Async.Task saveBlob(Container container, string v1, string v2, StorageType config);
-    Task<Guid> GetInstanceId();
+    public Async.Task SaveBlob(Container container, string v1, string v2, StorageType config);
+    public Async.Task<Guid> GetInstanceId();
+
+    public Async.Task<Uri?> GetFileUrl(Container container, string name, StorageType storageType);
+
+    public Async.Task<Uri?> GetContainerSasUrl(Container container, StorageType storageType, BlobSasPermissions permissions);
 }
 
+
 public class Containers : IContainers {
     private ILogTracer _log;
     private IStorage _storage;
     private ICreds _creds;
     private ArmClient _armClient;
+
     public Containers(ILogTracer log, IStorage storage, ICreds creds) {
         _log = log;
         _storage = storage;
         _creds = creds;
         _armClient = creds.ArmClient;
     }
-    public async Task<BinaryData?> GetBlob(Container container, string name, StorageType storageType) {
+
+    public async Async.Task<Uri?> GetFileUrl(Container container, string name, StorageType storageType) {
+        var client = await FindContainer(container, storageType);
+        if (client is null)
+            return null;
+
+        return new Uri($"{GetUrl(client.AccountName)}{container}/{name}");
+    }
+
+    public async Async.Task<BinaryData?> GetBlob(Container container, string name, StorageType storageType) {
         var client = await FindContainer(container, storageType);
 
         if (client == null) {
@@ -52,17 +67,23 @@ public Containers(ILogTracer log, IStorage storage, ICreds creds) {
         // #
         // # Secondary accounts, if they exist, are preferred for containers and have
         // # increased IOP rates, this should be a slight optimization
-        return await _storage.GetAccounts(storageType)
+
+        var containers = _storage.GetAccounts(storageType)
             .Reverse()
-            .Select(account => GetBlobService(account)?.GetBlobContainerClient(container.ContainerName))
-            .ToAsyncEnumerable()
-            .WhereAwait(async client => client != null && (await client.ExistsAsync()).Value)
-            .FirstOrDefaultAsync();
+            .Select(async account => (await GetBlobService(account))?.GetBlobContainerClient(container.ContainerName));
+
+        foreach (var c in containers) {
+            var client = await c;
+            if (client != null && (await client.ExistsAsync()).Value) {
+                return client;
+            }
+        }
+        return null;
     }
 
-    private BlobServiceClient? GetBlobService(string accountId) {
+    private async Async.Task<BlobServiceClient?> GetBlobService(string accountId) {
         _log.Info($"getting blob container (account_id: {accountId}");
-        var (accountName, accountKey) = _storage.GetStorageAccountNameAndKey(accountId);
+        var (accountName, accountKey) = await _storage.GetStorageAccountNameAndKey(accountId);
         if (accountName == null) {
             _log.Error("Failed to get storage account name");
             return null;
@@ -78,7 +99,7 @@ private static Uri GetUrl(string accountName) {
 
     public async Async.Task<Uri?> GetFileSasUrl(Container container, string name, StorageType storageType, BlobSasPermissions permissions, TimeSpan? duration = null) {
         var client = await FindContainer(container, storageType) ?? throw new Exception($"unable to find container: {container.ContainerName} - {storageType}");
-        var (accountName, accountKey) = _storage.GetStorageAccountNameAndKey(client.AccountName);
+        var (accountName, accountKey) = await _storage.GetStorageAccountNameAndKey(client.AccountName);
 
         var (startTime, endTime) = SasTimeWindow(duration ?? TimeSpan.FromDays(30));
 
@@ -108,17 +129,48 @@ private static Uri GetUrl(string accountName) {
         return (start, expiry);
     }
 
-    public async System.Threading.Tasks.Task saveBlob(Container container, string name, string data, StorageType storageType) {
+    public async Async.Task SaveBlob(Container container, string name, string data, StorageType storageType) {
         var client = await FindContainer(container, storageType) ?? throw new Exception($"unable to find container: {container.ContainerName} - {storageType}");
-
         await client.UploadBlobAsync(name, new BinaryData(data));
     }
 
+    //TODO: get this ones on startup and cache (and make this method un-accessible to everyone else)
     public async Async.Task<Guid> GetInstanceId() {
         var blob = await GetBlob(new Container("base-config"), "instance_id", StorageType.Config);
         if (blob == null) {
             throw new System.Exception("Blob Not Found");
         }
         return System.Guid.Parse(blob.ToString());
     }
-}
+
+    public Uri? GetContainerSasUrlService(
+        BlobContainerClient client,
+        BlobSasPermissions permissions,
+        bool tag = false,
+        TimeSpan? timeSpan = null) {
+        var (start, expiry) = SasTimeWindow(timeSpan ?? TimeSpan.FromDays(30.0));
+        var sasBuilder = new BlobSasBuilder(permissions, expiry) { StartsOn = start };
+        var sas = client.GenerateSasUri(sasBuilder);
+        return sas;
+    }
+
+
+    //TODO: instead of returning null when container not found, convert to return to "Result" type and set appropriate error
+    public async Async.Task<Uri?> GetContainerSasUrl(Container container, StorageType storageType, BlobSasPermissions permissions) {
+        var client = await FindContainer(container, storageType);
+
+        if (client is null) {
+            return null;
+        }
+
+        var uri = GetContainerSasUrlService(client, permissions);
+
+        if (uri is null) {
+            //TODO: return result error
+            return uri;
+        } else {
+            return uri;
+        }
+    }
+}
+

src/ApiService/ApiService/onefuzzlib/Creds.cs

@@ -8,7 +8,7 @@ namespace Microsoft.OneFuzz.Service;
 public interface ICreds {
     public DefaultAzureCredential GetIdentity();
 
-    public string GetSubcription();
+    public string GetSubscription();
 
     public string GetBaseResourceGroup();
 
@@ -21,6 +21,8 @@ public interface ICreds {
     public ResourceGroupResource GetResourceGroupResource();
 
     public string GetBaseRegion();
+
+    public Uri GetInstanceUrl();
 }
 
 public class Creds : ICreds {
@@ -33,14 +35,14 @@ public class Creds : ICreds {
     public Creds(IServiceConfig config) {
         _config = config;
         _azureCredential = new DefaultAzureCredential();
-        _armClient = new ArmClient(this.GetIdentity(), this.GetSubcription());
+        _armClient = new ArmClient(this.GetIdentity(), this.GetSubscription());
     }
 
     public DefaultAzureCredential GetIdentity() {
         return _azureCredential;
     }
 
-    public string GetSubcription() {
+    public string GetSubscription() {
         var storageResourceId = _config.OneFuzzDataStorage
             ?? throw new System.Exception("Data storage env var is not present");
         var storageResource = new ResourceIdentifier(storageResourceId);
@@ -75,4 +77,8 @@ public ResourceGroupResource GetResourceGroupResource() {
     public string GetBaseRegion() {
         return ArmClient.GetResourceGroupResource(GetResourceGroupResourceIdentifier()).Data.Location.Name;
     }
+
+    public Uri GetInstanceUrl() {
+        return new Uri($"https://{GetInstanceName()}.azurewebsites.net");
+    }
 }

src/ApiService/ApiService/onefuzzlib/IpOperations.cs

@@ -32,11 +32,11 @@ public async Async.Task<PublicIPAddressResource> GetIp(string resourceGroup, str
         return await _creds.GetResourceGroupResource().GetPublicIPAddressAsync(name);
     }
 
-    public System.Threading.Tasks.Task DeleteNic(string resourceGroup, string name) {
+    public Async.Task DeleteNic(string resourceGroup, string name) {
         throw new NotImplementedException();
     }
 
-    public System.Threading.Tasks.Task DeleteIp(string resourceGroup, string name) {
+    public Async.Task DeleteIp(string resourceGroup, string name) {
         throw new NotImplementedException();
     }
 }

src/ApiService/ApiService/onefuzzlib/JobOperations.cs

@@ -3,12 +3,12 @@
 namespace Microsoft.OneFuzz.Service;
 
 public interface IJobOperations : IStatefulOrm<Job, JobState> {
-    System.Threading.Tasks.Task<Job?> Get(Guid jobId);
-    System.Threading.Tasks.Task OnStart(Job job);
+    Async.Task<Job?> Get(Guid jobId);
+    Async.Task OnStart(Job job);
     IAsyncEnumerable<Job> SearchExpired();
-    System.Threading.Tasks.Task Stopping(Job job, ITaskOperations taskOperations);
+    Async.Task Stopping(Job job, ITaskOperations taskOperations);
     IAsyncEnumerable<Job> SearchState(IEnumerable<JobState> states);
-    System.Threading.Tasks.Task StopNeverStartedJobs();
+    Async.Task StopNeverStartedJobs();
 }
 
 public class JobOperations : StatefulOrm<Job, JobState>, IJobOperations {
@@ -18,11 +18,11 @@ public JobOperations(IStorage storage, ILogTracer logTracer, IServiceConfig conf
         _events = events;
     }
 
-    public async System.Threading.Tasks.Task<Job?> Get(Guid jobId) {
+    public async Async.Task<Job?> Get(Guid jobId) {
         return await QueryAsync($"PartitionKey eq '{jobId}'").FirstOrDefaultAsync();
     }
 
-    public async System.Threading.Tasks.Task OnStart(Job job) {
+    public async Async.Task OnStart(Job job) {
         if (job.EndTime == null) {
             await Replace(job with { EndTime = DateTimeOffset.UtcNow + TimeSpan.FromHours(job.Config.Duration) });
         }
@@ -40,11 +40,11 @@ public IAsyncEnumerable<Job> SearchState(IEnumerable<JobState> states) {
         return QueryAsync(filter: query);
     }
 
-    public System.Threading.Tasks.Task StopNeverStartedJobs() {
+    public Async.Task StopNeverStartedJobs() {
         throw new NotImplementedException();
     }
 
-    public async System.Threading.Tasks.Task Stopping(Job job, ITaskOperations taskOperations) {
+    public async Async.Task Stopping(Job job, ITaskOperations taskOperations) {
         job = job with { State = JobState.Stopping };
         var tasks = await taskOperations.QueryAsync(filter: $"job_id eq '{job.JobId}'").ToListAsync();
         var taskNotStopped = tasks.ToLookup(task => task.State != TaskState.Stopped);