From bd6f7b6d83a67f08100b3e3b2ccb49a83fc20861 Mon Sep 17 00:00:00 2001
From: Harshavardhana <harsha@minio.io>
Date: Sun, 20 Mar 2022 14:46:43 -0700
Subject: [PATCH] fix: make decommission restart non-blocking (#14591)

currently an on-going decommission, during a server
restart might block the startup sequence for relatively
longer periods, instead start the decommission in
background lazily.
---
 cmd/erasure-server-pool-decom.go | 17 ++++++++++-------
 cmd/erasure-server-pool.go       |  4 +++-
 2 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/cmd/erasure-server-pool-decom.go b/cmd/erasure-server-pool-decom.go
index 06c594015..acaf735b3 100644
--- a/cmd/erasure-server-pool-decom.go
+++ b/cmd/erasure-server-pool-decom.go
@@ -472,13 +472,16 @@ func (z *erasureServerPools) Init(ctx context.Context) error {
 		// '-1' as argument to decommission multiple pools at a time
 		// but this is not a priority at the moment.
 		for _, pool := range meta.returnResumablePools(1) {
-			err := z.Decommission(ctx, pool.ID)
-			switch err {
-			case errDecommissionAlreadyRunning:
-				fallthrough
-			case nil:
-				go z.doDecommissionInRoutine(ctx, pool.ID)
-			}
+			go func(pool PoolStatus) {
+				switch err := z.Decommission(ctx, pool.ID); err {
+				case errDecommissionAlreadyRunning:
+					fallthrough
+				case nil:
+					z.doDecommissionInRoutine(ctx, pool.ID)
+				default:
+					logger.LogIf(ctx, fmt.Errorf("Unable to resume decommission of pool %v: %w", pool, err))
+				}
+			}(pool)
 		}
 		z.poolMeta = meta
 
diff --git a/cmd/erasure-server-pool.go b/cmd/erasure-server-pool.go
index 10dc094e7..e427dc860 100644
--- a/cmd/erasure-server-pool.go
+++ b/cmd/erasure-server-pool.go
@@ -128,7 +128,9 @@ func newErasureServerPools(ctx context.Context, endpointServerPools EndpointServ
 			if !configRetriableErrors(err) {
 				logger.Fatal(err, "Unable to initialize backend")
 			}
-			time.Sleep(time.Duration(r.Float64() * float64(5*time.Second)))
+			retry := time.Duration(r.Float64() * float64(5*time.Second))
+			logger.LogIf(ctx, fmt.Errorf("Unable to initialize backend: %w, retrying in %s", err, retry))
+			time.Sleep(retry)
 			continue
 		}
 		break