- podman-etcd: set attributes if they fail during force-new-cluster
Resolves: RHEL-150700
This commit is contained in:
Oyvind Albrigtsen
parent
c33f29945a
commit
b841c79c29
@ -0,0 +1,111 @@
|
||||
From e4d311b40d8ded2a1921a0e5c01cb49a07c9fb35 Mon Sep 17 00:00:00 2001
|
||||
From: Carlo Lobrano <c.lobrano@gmail.com>
|
||||
Date: Thu, 5 Feb 2026 19:31:42 +0100
|
||||
Subject: [PATCH] podman-etcd: fix learner node attribute not set after etcdctl
|
||||
failure
|
||||
|
||||
Ensure that learner_node attribute is always set when the member list
|
||||
contains one learner member.
|
||||
|
||||
Moreover:
|
||||
* Ensure set_standalone_node is called after adding a learner member.
|
||||
* Capture stderr from etcdctl for better error logging.
|
||||
---
|
||||
heartbeat/podman-etcd | 61 +++++++++++++++++++++++++++----------------
|
||||
1 file changed, 38 insertions(+), 23 deletions(-)
|
||||
|
||||
diff --git a/heartbeat/podman-etcd b/heartbeat/podman-etcd
|
||||
index 77525ddb7..06814ad89 100755
|
||||
--- a/heartbeat/podman-etcd
|
||||
+++ b/heartbeat/podman-etcd
|
||||
@@ -1082,7 +1082,7 @@ add_member_as_learner()
|
||||
local peer_url=$(ip_url $member_ip)
|
||||
|
||||
ocf_log info "add $member_name ($member_ip) to the member list as learner"
|
||||
- out=$(podman exec "${CONTAINER}" etcdctl --endpoints="$endpoint_url:2379" member add "$member_name" --peer-urls="$peer_url:2380" --learner)
|
||||
+ out=$(podman exec "${CONTAINER}" etcdctl --endpoints="$endpoint_url:2379" member add "$member_name" --peer-urls="$peer_url:2380" --learner 2>&1)
|
||||
rc=$?
|
||||
if [ $rc -ne 0 ]; then
|
||||
ocf_log err "could not add $member_name as learner, error code $rc, etcdctl output: $out"
|
||||
@@ -1429,10 +1429,22 @@ detect_cluster_leadership_loss()
|
||||
manage_peer_membership()
|
||||
{
|
||||
local member_list_json="$1"
|
||||
+ local peer_ip_map_entry
|
||||
+ local peer_member_name
|
||||
+ local peer_member_ip
|
||||
+ local peer_member_id
|
||||
+
|
||||
+ # Get peer node name and IP
|
||||
+ peer_ip_map_entry=$(echo "$OCF_RESKEY_node_ip_map" | tr ';' '\n' | grep -vF "$NODENAME")
|
||||
+ if [ -z "$peer_ip_map_entry" ]; then
|
||||
+ ocf_exit_reason "manage_peer_membership: could not parse node_ip_map: '$OCF_RESKEY_node_ip_map'"
|
||||
+ exit $OCF_ERR_CONFIGURED
|
||||
+ fi
|
||||
+ peer_member_name=$(echo "$peer_ip_map_entry" | cut -d: -f1)
|
||||
+ peer_member_ip=$(echo "$peer_ip_map_entry" | cut -d: -f2-)
|
||||
|
||||
- # Example of .members[] instance fields in member list json format:
|
||||
- # NOTE that "name" is present in voting members only, while "isLearner" in learner members only
|
||||
- # and the value is always true (not a string) in that case.
|
||||
+ # Parsing the member list's json output to find a "learner" member.
|
||||
+ # Example of .members[] instance fields in member list json format:
|
||||
# {
|
||||
# "ID": <member ID>,
|
||||
# "name": "<node hostname>",
|
||||
@@ -1443,26 +1455,28 @@ manage_peer_membership()
|
||||
# "https://<node IP>:2379"
|
||||
# ]
|
||||
# }
|
||||
- for node in $(echo "$OCF_RESKEY_node_ip_map" | sed "s/\s//g;s/;/ /g"); do
|
||||
- name=$(echo "$node" | cut -d: -f1)
|
||||
- # do not check itself
|
||||
- if [ "$name" = "$NODENAME" ]; then
|
||||
- continue
|
||||
- fi
|
||||
+ # NOTE that the "name" field is present in voting members only, while "isLearner"
|
||||
+ # field in learner members only and the value is always true (not a string) in that case.
|
||||
+ peer_member_id=$(printf "%s" "$member_list_json" | jq -r ".members[] | select( .peerURLs | map(test(\"$peer_member_ip\")) | any).ID")
|
||||
+ if [ -z "$peer_member_id" ]; then
|
||||
+ ocf_log info "$peer_member_name is not in the members list"
|
||||
+ add_member_as_learner "$peer_member_name" "$peer_member_ip"
|
||||
+ set_standalone_node
|
||||
+ return
|
||||
+ fi
|
||||
|
||||
- # Check by IP instead of Name since "learner" members appear only in peerURLs, not by Name.
|
||||
- ip=$(echo "$node" | cut -d: -f2-) # Grab everything after the first : this covers ipv4/ipv6
|
||||
- peer_member_id=$(printf "%s" "$member_list_json" | jq -r ".members[] | select( .peerURLs | map(test(\"$ip\")) | any).ID")
|
||||
- if [ -z "$peer_member_id" ]; then
|
||||
- ocf_log info "$name is not in the members list"
|
||||
- add_member_as_learner "$name" "$ip"
|
||||
- set_standalone_node
|
||||
- else
|
||||
- ocf_log debug "$name is in the members list by IP: $ip"
|
||||
- # Errors from reconcile_member_state are logged internally. Ignoring them here prevents stopping a healthy voter agent; critical local failures are caught by detect_cluster_leadership_loss.
|
||||
- reconcile_member_state "$member_list_json"
|
||||
- fi
|
||||
- done
|
||||
+ # Ensure learner_node attribute is always set when we have a learner member
|
||||
+ local learner_member_id=$(printf "%s" "$member_list_json" | jq -r ".members[] | select( .isLearner==true ).ID")
|
||||
+ local current_learner_node=$(attribute_learner_node get)
|
||||
+ if [ -n "$learner_member_id" ] && [ -z "$current_learner_node" ]; then
|
||||
+ ocf_log debug "$peer_member_name found as learner in member list, but learner_node attribute was not set. Updating"
|
||||
+ attribute_learner_node update "$peer_member_name"
|
||||
+ return
|
||||
+ fi
|
||||
+
|
||||
+ ocf_log debug "$peer_member_name is in the members list by IP: $peer_member_ip"
|
||||
+ # Errors from reconcile_member_state are logged internally. Ignoring them here prevents stopping a healthy voter agent; critical local failures are caught by detect_cluster_leadership_loss.
|
||||
+ reconcile_member_state "$member_list_json"
|
||||
}
|
||||
|
||||
check_peer()
|
||||
@@ -2209,6 +2223,7 @@ podman_start()
|
||||
peer_node_ip="$(attribute_node_ip_peer)"
|
||||
if [ -n "$peer_node_name" ] && [ -n "$peer_node_ip" ]; then
|
||||
add_member_as_learner "$peer_node_name" "$peer_node_ip"
|
||||
+ set_standalone_node
|
||||
else
|
||||
ocf_log err "could not add peer as learner (peer node name: ${peer_node_name:-unknown}, peer ip: ${peer_node_ip:-unknown})"
|
||||
fi
|
||||
@ -45,7 +45,7 @@
|
||||
Name: resource-agents
|
||||
Summary: Open Source HA Reusable Cluster Resource Scripts
|
||||
Version: 4.10.0
|
||||
Release: 107%{?rcver:%{rcver}}%{?numcomm:.%{numcomm}}%{?alphatag:.%{alphatag}}%{?dirty:.%{dirty}}%{?dist}
|
||||
Release: 108%{?rcver:%{rcver}}%{?numcomm:.%{numcomm}}%{?alphatag:.%{alphatag}}%{?dirty:.%{dirty}}%{?dist}
|
||||
License: GPLv2+ and LGPLv2+
|
||||
URL: https://github.com/ClusterLabs/resource-agents
|
||||
Source0: %{upstream_prefix}-%{upstream_version}.tar.gz
|
||||
@ -198,6 +198,7 @@ Patch145: RHEL-139519-podman-etcd-verify-no-containers-running-or-being-deleted.
|
||||
Patch146: RHEL-42513-powervs-subnet-wait-for-IP.patch
|
||||
Patch147: RHEL-143527-powervs-move-ip-powervs-subnet-fix-error-logging.patch
|
||||
Patch148: RHEL-145628-podman-etcd-enhance-etcd-data-backup-with-snapshots-and-retention.patch
|
||||
Patch149: RHEL-150700-podman-etcd-set-attributes-if-they-fail-during-force-new-cluster.patch
|
||||
|
||||
# bundled ha-cloud-support libs
|
||||
Patch500: ha-cloud-support-aliyun.patch
|
||||
@ -491,6 +492,7 @@ exit 1
|
||||
%patch -p1 -P 146
|
||||
%patch -p1 -P 147
|
||||
%patch -p1 -P 148
|
||||
%patch -p1 -P 149
|
||||
|
||||
# bundled ha-cloud-support libs
|
||||
%patch -p1 -P 500
|
||||
@ -823,6 +825,11 @@ rm -rf %{buildroot}/usr/share/doc/resource-agents
|
||||
%{_usr}/lib/ocf/lib/heartbeat/OCF_*.pm
|
||||
|
||||
%changelog
|
||||
* Wed Feb 18 2026 Oyvind Albrigtsen <oalbrigt@redhat.com> - 4.10.0-108
|
||||
- podman-etcd: set attributes if they fail during force-new-cluster
|
||||
|
||||
Resolves: RHEL-150700
|
||||
|
||||
* Wed Feb 4 2026 Oyvind Albrigtsen <oalbrigt@redhat.com> - 4.10.0-107
|
||||
- podman-etcd: enhance etcd data backup with snapshots and retention
|
||||
|
||||
|
||||
Loading…
Reference in New Issue
Block a user