test: Divergent test 3 with pg_autoscale_mode on pick divergent osd

When creating new pool, the current code pick the divergent osd by
the first pg out of pg dump pgs, that can be in "unknown" status
which means the up_primary = -1 and that will fail the test.
We need to wait unitl the first pg is active+clean

Fixes: https://tracker.ceph.com/issues/56034
Signed-off-by: Nitzan Mordechai <nmordech@redhat.com>
(cherry picked from commit 13c640b5a82bf8f9b1aa5619416a4dbbb5d9a8d0)

diff --git a/qa/standalone/osd/divergent-priors.sh b/qa/standalone/osd/divergent-priors.sh
index dec0e7ad4e1bca0d048fbbf4495bf309a1fcf804..40d72544d1b062b1b17f0677ccd8f25bef3ffe59 100755 (executable)
--- a/qa/standalone/osd/divergent-priors.sh
+++ b/qa/standalone/osd/divergent-priors.sh
@@ -650,13 +650,28 @@ function TEST_divergent_3() {
     # reproduce https://tracker.ceph.com/issues/41816
     ceph osd pool set $poolname pg_autoscale_mode on
 
-    flush_pg_stats || return 1
-    wait_for_clean || return 1
+    divergent=-1
+    start_time=$(date +%s)
+    max_duration=300
+
+    while [ "$divergent" -le -1 ]
+      do
+        flush_pg_stats || return 1
+        wait_for_clean || return 1
+
+        # determine primary
+        divergent="$(ceph pg dump pgs --format=json | jq '.pg_stats[0].up_primary')"
+        echo "primary and soon to be divergent is $divergent"
+        ceph pg dump pgs
+
+        current_time=$(date +%s)
+        elapsed_time=$(expr $current_time - $start_time)
+        if [ "$elapsed_time" -gt "$max_duration" ]; then
+          echo "timed out waiting for divergent"
+          return 1
+        fi
+    done
 
-    # determine primary
-    local divergent="$(ceph pg dump pgs --format=json | jq '.pg_stats[0].up_primary')"
-    echo "primary and soon to be divergent is $divergent"
-    ceph pg dump pgs
     local non_divergent=""
     for i in $osds
     do