Merge #1614

1614: fix(nvmf): fixing duplicate QID error r=dsavitskiy a=dsavitskiy

* Updating to the SPDK with the fix
* A test for many parallel rebuilds added

Co-authored-by: Dmitry Savitskiy <[email protected]>

io-engine/tests/nexus_rebuild_parallel.rs

@@ -0,0 +1,207 @@
+pub mod common;
+
+use std::time::{Duration, Instant};
+
+use common::{
+    compose::{
+        rpc::v1::{nexus::ChildState, GrpcConnect, Status},
+        Binary,
+        Builder,
+    },
+    nexus::NexusBuilder,
+    pool::PoolBuilder,
+    replica::ReplicaBuilder,
+};
+use tonic::Code;
+
+struct Volume {
+    replicas: Vec<ReplicaBuilder>,
+    nex: NexusBuilder,
+}
+
+/// Runs multiple rebuild jobs parallel.
+#[tokio::test]
+async fn nexus_rebuild_parallel() {
+    common::composer_init();
+
+    const R: usize = 3; // Number of replicas / nodes.
+    const N: usize = 20; // Number f volumes
+    const REPL_SIZE: u64 = 50;
+    const NEXUS_SIZE: u64 = REPL_SIZE;
+    const POOL_SIZE: u64 = REPL_SIZE * N as u64 + REPL_SIZE;
+    const DISK_SIZE: u64 = POOL_SIZE * 1024 * 1024;
+    const DISK_NAME: &str = "disk";
+
+    // Create pool data file.
+    for r in 0 .. R {
+        let name = format!("/tmp/{DISK_NAME}_{r}");
+        common::delete_file(&[name.clone()]);
+        common::truncate_file_bytes(&name, DISK_SIZE);
+    }
+
+    let test = Builder::new()
+        .name("cargo-test")
+        .network("10.1.0.0/16")
+        .unwrap()
+        .add_container_bin(
+            "ms_0",
+            Binary::from_dbg("io-engine")
+                .with_args(vec!["-l", "1,2", "-Fcolor,compact,nodate,host"])
+                .with_bind("/tmp", "/host/tmp"),
+        )
+        .add_container_bin(
+            "ms_1",
+            Binary::from_dbg("io-engine")
+                .with_args(vec!["-l", "2,3", "-Fcolor,compact,nodate,host"])
+                .with_bind("/tmp", "/host/tmp"),
+        )
+        .add_container_bin(
+            "ms_2",
+            Binary::from_dbg("io-engine")
+                .with_args(vec!["-l", "4,5", "-Fcolor,compact,nodate,host"])
+                .with_bind("/tmp", "/host/tmp"),
+        )
+        .with_clean(true)
+        .build()
+        .await
+        .unwrap();
+
+    let conn = GrpcConnect::new(&test);
+
+    let ms_nex = conn.grpc_handle_shared("ms_2").await.unwrap();
+
+    let ms = [
+        conn.grpc_handle_shared("ms_0").await.unwrap(),
+        conn.grpc_handle_shared("ms_1").await.unwrap(),
+        conn.grpc_handle_shared("ms_2").await.unwrap(),
+    ];
+
+    let mut pools = Vec::new();
+
+    // Create pools.
+    for (r, ms) in ms.iter().enumerate() {
+        let bdev = format!("aio:///host/tmp/{DISK_NAME}_{r}?blk_size=512");
+        let mut pool = PoolBuilder::new(ms.clone())
+            .with_name(&format!("p{r}"))
+            .with_new_uuid()
+            .with_bdev(&bdev);
+        pool.create().await.unwrap();
+        pools.push(pool);
+    }
+
+    let mut vols = Vec::new();
+
+    for i in 0 .. N {
+        // Create R replicas on the pools.
+        let mut replicas = Vec::new();
+        for r in 0 .. R {
+            let mut repl = ReplicaBuilder::new(ms[r].clone())
+                .with_pool(&pools[r])
+                .with_name(&format!("v{i}r{r}"))
+                .with_new_uuid()
+                .with_size_mb(REPL_SIZE)
+                .with_thin(false);
+            repl.create().await.unwrap();
+            repl.share().await.unwrap();
+            replicas.push(repl);
+        }
+
+        // Create the nexus with 2 replicas @ 1,2.
+        let mut nex = NexusBuilder::new(ms_nex.clone())
+            .with_name(&format!("v{i}"))
+            .with_new_uuid()
+            .with_size_mb(NEXUS_SIZE)
+            .with_replica(&replicas[1])
+            .with_replica(&replicas[2]);
+
+        nex.create().await.unwrap();
+
+        vols.push(Volume {
+            replicas,
+            nex,
+        });
+    }
+
+    // Adding replicas / starting rebuilds.
+    for vol in &vols {
+        vol.nex.add_replica(&vol.replicas[0], false).await.unwrap();
+    }
+
+    monitor_volumes(&vols, Duration::from_secs(30))
+        .await
+        .expect("All volumes must go online");
+
+    // Delete test files.
+    for r in 0 .. R {
+        let name = format!("/tmp/{DISK_NAME}_{r}");
+        common::delete_file(&[name.clone()]);
+    }
+}
+
+/// Monitors and prints volume states.
+async fn monitor_volumes(
+    vols: &Vec<Volume>,
+    timeout: Duration,
+) -> Result<(), Status> {
+    println!("\nMonitoring {n} volumes", n = vols.len());
+
+    let start = Instant::now();
+
+    loop {
+        let mut vols_degraded = false;
+        let mut vols_failed = false;
+
+        for vol in vols {
+            let mut s = String::new();
+
+            let n = vol.nex.get_nexus().await.unwrap();
+            for c in n.children.iter() {
+                match c.state() {
+                    ChildState::Faulted => {
+                        s = format!("{s} FAILED     | ");
+                        vols_failed = true;
+                    }
+                    ChildState::Online => {
+                        s = format!("{s} ONLINE     | ");
+                    }
+                    ChildState::Unknown => {
+                        s = format!("{s} UNKNOWN    | ");
+                        vols_degraded = true;
+                    }
+                    ChildState::Degraded => {
+                        s = format!(
+                            "{s} REBUILD {p:02} | ",
+                            p = c.rebuild_progress
+                        );
+                        vols_degraded = true;
+                    }
+                }
+            }
+
+            println!("    {n}: {s}", n = n.name)
+        }
+        println!("-");
+
+        if vols_failed {
+            println!("One or more volumes failed");
+            return Err(Status::new(
+                Code::Internal,
+                "One or more volumes failed".to_string(),
+            ));
+        }
+
+        if !vols_degraded {
+            println!("All volumes are online");
+            return Ok(());
+        }
+
+        if start.elapsed() > timeout {
+            return Err(Status::new(
+                Code::Cancelled,
+                "Waiting for volumes to go Online state timed out".to_string(),
+            ));
+        }
+
+        tokio::time::sleep(std::time::Duration::from_millis(1000)).await;
+    }
+}

nix/pkgs/libspdk/default.nix

@@ -56,13 +56,13 @@ let
   # 7. Copy SHA256 from 'got' of the error message to 'sha256' field.
   # 8. 'nix-shell' build must now succeed.
   drvAttrs = rec {
-    version = "23.05-baffd90";
+    version = "23.05-759538c";
 
     src = fetchFromGitHub {
       owner = "openebs";
       repo = "spdk";
-      rev = "baffd90809bdd0b113b76fc7c9d7663b69d26752";
-      sha256 = "sha256-tyxtXh7RpU6VtBlEjZ5MotnKQ4uZbbLD5sV+ndkuHhc=";
+      rev = "759538c28701743c1db23867deb3094819943250";
+      sha256 = "sha256-m4zZLd8FTosSGUS0l8cJG69CtmCL1Blx5SgvkH7qao4=";
       fetchSubmodules = true;
     };
 

test/python/requirements.txt

@@ -2,6 +2,7 @@ asyncio==3.4.3
 asyncssh==2.14.1
 black==22.10.0
 protobuf==4.21.8
+pytest==7.4.4
 pytest-asyncio==0.21.1
 pytest-bdd==6.1.1
 pytest-black==0.3.12