DAOS-16477 pool: Rename Suspect state to Dead (#15584)

Change the name to more closely reflect the underlying SWIM status, and reduce user confusion. An engine that has been marked DEAD by SWIM cannot participate in pool services, and has most likely already SIGKILL-ed itself. Signed-off-by: Michael MacDonald <[email protected]>
daos-stack · Dec 13, 2024 · 29ad64d · 29ad64d
1 parent 1392ea8
commit 29ad64d
Show file tree

Hide file tree

Showing 24 changed files with 363 additions and 500 deletions.
diff --git a/docs/admin/pool_operations.md b/docs/admin/pool_operations.md
@@ -698,25 +698,25 @@ The example below shows a rebuild in progress and NVMe space allocated.
     Rebuild busy, 75 objs, 9722 recs
 ```
 
-After experiencing significant failures, the pool may retain some suspect
+After experiencing significant failures, the pool may retain some "dead"
 engines that have been marked as DEAD by the SWIM protocol but were not excluded
 from the pool to prevent potential data inconsistency. An administrator can bring
 these engines back online by restarting them. The example below illustrates the
-system’s status with suspect and disabled engines.
+system’s status with dead and disabled engines.
 
 ```bash
 $ dmg pool query tank -t
 ```
 
 NB: The --health-only/-t option is necessary to conduct pool health-related queries only.
-This is important because suspect ranks may cause commands to hang and timeout so identifying
+This is important because dead ranks may cause commands to hang and timeout so identifying
 and restarting them is a useful procedure.
 
 ```bash
 Pool 6f450a68-8c7d-4da9-8900-02691650f6a2, ntarget=8, disabled=2, leader=3, version=4, state=Degraded
     Pool health info:
     - Disabled ranks: 1
-    - Suspect ranks: 2
+    - Dead ranks: 2
     - Rebuild busy, 0 objs, 0 recs
 ```
 

diff --git a/src/control/cmd/daos/health.go b/src/control/cmd/daos/health.go
@@ -100,7 +100,7 @@ func (cmd *healthCheckCmd) Execute([]string) error {
 		}()
 
 		queryMask := daos.MustNewPoolQueryMask(daos.PoolQueryOptionEnabledEngines,
-			daos.PoolQueryOptionSuspectEngines)
+			daos.PoolQueryOptionDeadEngines)
 		if pool.DisabledTargets > 0 {
 			queryMask.SetOptions(daos.PoolQueryOptionDisabledEngines)
 		}
@@ -111,7 +111,7 @@ func (cmd *healthCheckCmd) Execute([]string) error {
 		}
 		pool.EnabledRanks = tpi.EnabledRanks
 		pool.DisabledRanks = tpi.DisabledRanks
-		pool.SuspectRanks = tpi.SuspectRanks
+		pool.DeadRanks = tpi.DeadRanks
 
 		poolConts, err := listContainers(poolHdl)
 		if err != nil {

diff --git a/src/control/cmd/daos/pool.go b/src/control/cmd/daos/pool.go
@@ -300,7 +300,7 @@ func queryPoolRankLists(poolHdl C.daos_handle_t, queryMask daos.PoolQueryMask) (
 	var rl *C.d_rank_list_t = nil
 
 	if queryMask.HasOption(daos.PoolQueryOptionEnabledEngines) || queryMask.HasOption(daos.PoolQueryOptionDisabledEngines) ||
-		queryMask.HasOption(daos.PoolQueryOptionSuspectEngines) {
+		queryMask.HasOption(daos.PoolQueryOptionDeadEngines) {
 		rlPtr = &rl
 	}
 
@@ -330,8 +330,8 @@ func queryPoolRankLists(poolHdl C.daos_handle_t, queryMask daos.PoolQueryMask) (
 		if queryMask.HasOption(daos.PoolQueryOptionDisabledEngines) {
 			poolInfo.DisabledRanks = rs
 		}
-		if queryMask.HasOption(daos.PoolQueryOptionSuspectEngines) {
-			poolInfo.SuspectRanks = rs
+		if queryMask.HasOption(daos.PoolQueryOptionDeadEngines) {
+			poolInfo.DeadRanks = rs
 		}
 	}
 
@@ -357,8 +357,8 @@ func queryPool(poolHdl C.daos_handle_t, queryMask daos.PoolQueryMask) (*daos.Poo
 			poolInfo.EnabledRanks = poolInfo1.EnabledRanks
 		case daos.PoolQueryOptionDisabledEngines:
 			poolInfo.DisabledRanks = poolInfo1.DisabledRanks
-		case daos.PoolQueryOptionSuspectEngines:
-			poolInfo.SuspectRanks = poolInfo1.SuspectRanks
+		case daos.PoolQueryOptionDeadEngines:
+			poolInfo.DeadRanks = poolInfo1.DeadRanks
 		}
 		return nil
 	}
@@ -369,8 +369,8 @@ func queryPool(poolHdl C.daos_handle_t, queryMask daos.PoolQueryMask) (*daos.Poo
 		firstOption = daos.PoolQueryOptionEnabledEngines
 	} else if originalMask.HasOption(daos.PoolQueryOptionDisabledEngines) {
 		firstOption = daos.PoolQueryOptionDisabledEngines
-	} else if originalMask.HasOption(daos.PoolQueryOptionSuspectEngines) {
-		firstOption = daos.PoolQueryOptionSuspectEngines
+	} else if originalMask.HasOption(daos.PoolQueryOptionDeadEngines) {
+		firstOption = daos.PoolQueryOptionDeadEngines
 	}
 
 	// Perform the first query to get basic information
@@ -382,7 +382,7 @@ func queryPool(poolHdl C.daos_handle_t, queryMask daos.PoolQueryMask) (*daos.Poo
 	queryOptions := []string{
 		daos.PoolQueryOptionEnabledEngines,
 		daos.PoolQueryOptionDisabledEngines,
-		daos.PoolQueryOptionSuspectEngines,
+		daos.PoolQueryOptionDeadEngines,
 	}
 
 	// Process each option sequentially

diff --git a/src/control/cmd/daos/pretty/health.go b/src/control/cmd/daos/pretty/health.go
@@ -61,12 +61,12 @@ func printPoolHealth(out io.Writer, pi *daos.PoolInfo, verbose bool) {
 	}
 
 	var healthStrings []string
-	if pi.SuspectRanks != nil && pi.SuspectRanks.Count() > 0 {
-		degStr := "Suspect"
+	if pi.DeadRanks != nil && pi.DeadRanks.Count() > 0 {
+		deadStr := "Dead"
 		if verbose {
-			degStr += fmt.Sprintf(" %s", pi.SuspectRanks)
+			deadStr += fmt.Sprintf(" %s", pi.DeadRanks)
 		}
-		healthStrings = append(healthStrings, degStr)
+		healthStrings = append(healthStrings, deadStr)
 	}
 	if pi.DisabledTargets > 0 {
 		degStr := "Degraded"

diff --git a/src/control/cmd/daos/pretty/pool.go b/src/control/cmd/daos/pretty/pool.go
@@ -76,9 +76,9 @@ func PrintPoolInfo(pi *daos.PoolInfo, out io.Writer) error {
 	if pi.DisabledRanks.Count() > 0 {
 		fmt.Fprintf(w, "- Disabled ranks: %s\n", pi.DisabledRanks)
 	}
-	if pi.QueryMask.HasOption(daos.PoolQueryOptionSuspectEngines) &&
-		pi.SuspectRanks != nil && pi.SuspectRanks.Count() > 0 {
-		fmt.Fprintf(w, "- Suspect ranks: %s\n", pi.SuspectRanks)
+	if pi.QueryMask.HasOption(daos.PoolQueryOptionDeadEngines) &&
+		pi.DeadRanks != nil && pi.DeadRanks.Count() > 0 {
+		fmt.Fprintf(w, "- Dead ranks: %s\n", pi.DeadRanks)
 	}
 	if pi.Rebuild != nil {
 		if pi.Rebuild.Status == 0 {

diff --git a/src/control/cmd/daos/pretty/pool_test.go b/src/control/cmd/daos/pretty/pool_test.go
@@ -128,7 +128,7 @@ Pool space info:
   Free: 1 B, min:0 B, max:0 B, mean:0 B
 `, poolUUID.String()),
 		},
-		"normal response; suspect ranks": {
+		"normal response; dead ranks": {
 			pi: &daos.PoolInfo{
 				QueryMask:        daos.HealthOnlyPoolQueryMask,
 				State:            daos.PoolServiceStateDegraded,
@@ -141,7 +141,7 @@ Pool space info:
 				PoolLayoutVer:    1,
 				UpgradeLayoutVer: 2,
 				DisabledRanks:    ranklist.MustCreateRankSet("[0,1,3]"),
-				SuspectRanks:     ranklist.MustCreateRankSet("[2]"),
+				DeadRanks:        ranklist.MustCreateRankSet("[2]"),
 				Rebuild: &daos.PoolRebuildStatus{
 					State:   daos.PoolRebuildStateBusy,
 					Objects: 42,
@@ -163,7 +163,7 @@ Pool %s, ntarget=2, disabled=1, leader=42, version=100, state=Degraded
 Pool layout out of date (1 < 2) -- see `+backtickStr+` for details.
 Pool health info:
 - Disabled ranks: 0-1,3
-- Suspect ranks: 2
+- Dead ranks: 2
 - Rebuild busy, 42 objs, 21 recs
 `, poolUUID.String()),
 		},