.gitignore

@@ -1,2 +1,2 @@
-SOURCES/nagios-agents-metadata-105ab8a.tar.gz
-SOURCES/pacemaker-6fdc9deea.tar.gz
+SOURCES/nagios-agents-metadata-105ab8a7b2c16b9a29cf1c1596b80136eeef332b.tar.gz
+SOURCES/pacemaker-a3f44794f.tar.gz

.pacemaker.metadata

@@ -1,2 +1,2 @@
-ea6c0a27fd0ae8ce02f84a11f08a0d79377041c3 SOURCES/nagios-agents-metadata-105ab8a.tar.gz
-fbf71fb3fb42c76f9f1e98497505eb8521cab55e SOURCES/pacemaker-6fdc9deea.tar.gz
+2cbec94ad67dfbeba75e38d2c3c5c44961b3cd16 SOURCES/nagios-agents-metadata-105ab8a7b2c16b9a29cf1c1596b80136eeef332b.tar.gz
+b16198db5f86857ba8bc0ebd04fd386da360478a SOURCES/pacemaker-a3f44794f.tar.gz

SOURCES/001-remote-start-state.patch

@@ -1,402 +0,0 @@
-From cf53f523e691295879cd75cff1a86bc15664fa51 Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 09:59:13 -0400
-Subject: [PATCH 1/7] Feature: daemons: Add start state to LRMD handshake XML
-
-This gets read out of /etc/sysconfig/pacemaker and set into the
-environment.  The remote node executor will then add that to the XML
-that it sends to the controller upon startup.
-
-Ref T183
----
- daemons/execd/execd_commands.c | 5 +++++
- include/crm_internal.h         | 1 +
- 2 files changed, 6 insertions(+)
-
-diff --git a/daemons/execd/execd_commands.c b/daemons/execd/execd_commands.c
-index fa2761e..9a783a5 100644
---- a/daemons/execd/execd_commands.c
-+++ b/daemons/execd/execd_commands.c
-@@ -1474,6 +1474,7 @@ process_lrmd_signon(pcmk__client_t *client, xmlNode *request, int call_id,
-     int rc = pcmk_ok;
-     time_t now = time(NULL);
-     const char *protocol_version = crm_element_value(request, F_LRMD_PROTOCOL_VERSION);
-+    const char *start_state = pcmk__env_option(PCMK__ENV_NODE_START_STATE);
- 
-     if (compare_version(protocol_version, LRMD_MIN_PROTOCOL_VERSION) < 0) {
-         crm_err("Cluster API version must be greater than or equal to %s, not %s",
-@@ -1503,6 +1504,10 @@ process_lrmd_signon(pcmk__client_t *client, xmlNode *request, int call_id,
-     crm_xml_add(*reply, F_LRMD_PROTOCOL_VERSION, LRMD_PROTOCOL_VERSION);
-     crm_xml_add_ll(*reply, PCMK__XA_UPTIME, now - start_time);
- 
-+    if (start_state) {
-+        crm_xml_add(*reply, PCMK__XA_NODE_START_STATE, start_state);
-+    }
-+
-     return rc;
- }
- 
-diff --git a/include/crm_internal.h b/include/crm_internal.h
-index 5f6531f..771bd26 100644
---- a/include/crm_internal.h
-+++ b/include/crm_internal.h
-@@ -84,6 +84,7 @@
- #define PCMK__XA_GRAPH_ERRORS           "graph-errors"
- #define PCMK__XA_GRAPH_WARNINGS         "graph-warnings"
- #define PCMK__XA_MODE                   "mode"
-+#define PCMK__XA_NODE_START_STATE       "node_start_state"
- #define PCMK__XA_TASK                   "task"
- #define PCMK__XA_UPTIME                 "uptime"
- #define PCMK__XA_CONN_HOST              "connection_host"
--- 
-2.31.1
-
-From c950291742711b5c4c8986adc8e938fe6fef861c Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 10:04:32 -0400
-Subject: [PATCH 2/7] Feature: liblrmd: Save a remote node's requested start
- state
-
-Ref T183
----
- include/crm/common/ipc_internal.h | 1 +
- lib/lrmd/lrmd_client.c            | 7 +++++++
- 2 files changed, 8 insertions(+)
-
-diff --git a/include/crm/common/ipc_internal.h b/include/crm/common/ipc_internal.h
-index 5099dda..d203924 100644
---- a/include/crm/common/ipc_internal.h
-+++ b/include/crm/common/ipc_internal.h
-@@ -112,6 +112,7 @@ struct pcmk__remote_s {
-     int tcp_socket;
-     mainloop_io_t *source;
-     time_t uptime;
-+    char *start_state;
- 
-     /* CIB-only */
-     char *token;
-diff --git a/lib/lrmd/lrmd_client.c b/lib/lrmd/lrmd_client.c
-index c565728..4239105 100644
---- a/lib/lrmd/lrmd_client.c
-+++ b/lib/lrmd/lrmd_client.c
-@@ -588,7 +588,9 @@ lrmd_tls_connection_destroy(gpointer userdata)
-     }
- 
-     free(native->remote->buffer);
-+    free(native->remote->start_state);
-     native->remote->buffer = NULL;
-+    native->remote->start_state = NULL;
-     native->source = 0;
-     native->sock = 0;
-     native->psk_cred_c = NULL;
-@@ -980,6 +982,7 @@ lrmd_handshake(lrmd_t * lrmd, const char *name)
-         const char *version = crm_element_value(reply, F_LRMD_PROTOCOL_VERSION);
-         const char *msg_type = crm_element_value(reply, F_LRMD_OPERATION);
-         const char *tmp_ticket = crm_element_value(reply, F_LRMD_CLIENTID);
-+        const char *start_state = crm_element_value(reply, PCMK__XA_NODE_START_STATE);
-         long long uptime = -1;
- 
-         crm_element_value_int(reply, F_LRMD_RC, &rc);
-@@ -992,6 +995,10 @@ lrmd_handshake(lrmd_t * lrmd, const char *name)
-         crm_element_value_ll(reply, PCMK__XA_UPTIME, &uptime);
-         native->remote->uptime = uptime;
- 
-+        if (start_state) {
-+            native->remote->start_state = strdup(start_state);
-+        }
-+
-         if (rc == -EPROTO) {
-             crm_err("Executor protocol version mismatch between client (%s) and server (%s)",
-                 LRMD_PROTOCOL_VERSION, version);
--- 
-2.31.1
-
-From 7302014c7b7296be31b1f542b3f107d55b1fb2a0 Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 10:05:13 -0400
-Subject: [PATCH 3/7] Feature: liblrmd: Add lrmd__node_start_state.
-
-This function is used to get the start state out of an lrmd_private_t
-structure.
-
-Ref T183
----
- include/crm/lrmd_internal.h |  1 +
- lib/lrmd/lrmd_client.c      | 12 ++++++++++++
- 2 files changed, 13 insertions(+)
-
-diff --git a/include/crm/lrmd_internal.h b/include/crm/lrmd_internal.h
-index 5810554..d1cd25d 100644
---- a/include/crm/lrmd_internal.h
-+++ b/include/crm/lrmd_internal.h
-@@ -47,6 +47,7 @@ void lrmd__set_result(lrmd_event_data_t *event, enum ocf_exitcode rc,
- void lrmd__reset_result(lrmd_event_data_t *event);
- 
- time_t lrmd__uptime(lrmd_t *lrmd);
-+const char *lrmd__node_start_state(lrmd_t *lrmd);
- 
- /* Shared functions for IPC proxy back end */
- 
-diff --git a/lib/lrmd/lrmd_client.c b/lib/lrmd/lrmd_client.c
-index 4239105..82434b9 100644
---- a/lib/lrmd/lrmd_client.c
-+++ b/lib/lrmd/lrmd_client.c
-@@ -2538,3 +2538,15 @@ lrmd__uptime(lrmd_t *lrmd)
-         return native->remote->uptime;
-     }
- }
-+
-+const char *
-+lrmd__node_start_state(lrmd_t *lrmd)
-+{
-+    lrmd_private_t *native = lrmd->lrmd_private;
-+
-+    if (native->remote == NULL) {
-+        return NULL;
-+    } else {
-+        return native->remote->start_state;
-+    }
-+}
--- 
-2.31.1
-
-From e5e4d43f847da0930bae12f63c7e9d9c44c07cdf Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 10:07:58 -0400
-Subject: [PATCH 4/7] Refactor: controller: Make set_join_state a public
- function.
-
-This already does all the work of setting a node's start state.  It just
-needs to be made public and given arguments for what node to set instead
-of reading globals.
-
-Ref T183
----
- daemons/controld/controld_join_client.c | 20 ++++++++++----------
- daemons/controld/pacemaker-controld.h   |  3 +++
- 2 files changed, 13 insertions(+), 10 deletions(-)
-
-diff --git a/daemons/controld/controld_join_client.c b/daemons/controld/controld_join_client.c
-index da6a9d6..07e2a27 100644
---- a/daemons/controld/controld_join_client.c
-+++ b/daemons/controld/controld_join_client.c
-@@ -195,32 +195,31 @@ join_query_callback(xmlNode * msg, int call_id, int rc, xmlNode * output, void *
-     free_xml(generation);
- }
- 
--static void
--set_join_state(const char * start_state)
-+void
-+set_join_state(const char *start_state, const char *node_name, const char *node_uuid)
- {
-     if (pcmk__str_eq(start_state, "standby", pcmk__str_casei)) {
-         crm_notice("Forcing node %s to join in %s state per configured "
--                   "environment", controld_globals.our_nodename, start_state);
-+                   "environment", node_name, start_state);
-         cib__update_node_attr(controld_globals.logger_out,
-                               controld_globals.cib_conn, cib_sync_call,
--                              XML_CIB_TAG_NODES, controld_globals.our_uuid,
-+                              XML_CIB_TAG_NODES, node_uuid,
-                               NULL, NULL, NULL, "standby", "on", NULL, NULL);
- 
-     } else if (pcmk__str_eq(start_state, "online", pcmk__str_casei)) {
-         crm_notice("Forcing node %s to join in %s state per configured "
--                   "environment", controld_globals.our_nodename, start_state);
-+                   "environment", node_name, start_state);
-         cib__update_node_attr(controld_globals.logger_out,
-                               controld_globals.cib_conn, cib_sync_call,
--                              XML_CIB_TAG_NODES, controld_globals.our_uuid,
-+                              XML_CIB_TAG_NODES, node_uuid,
-                               NULL, NULL, NULL, "standby", "off", NULL, NULL);
- 
-     } else if (pcmk__str_eq(start_state, "default", pcmk__str_casei)) {
--        crm_debug("Not forcing a starting state on node %s",
--                  controld_globals.our_nodename);
-+        crm_debug("Not forcing a starting state on node %s", node_name);
- 
-     } else {
-         crm_warn("Unrecognized start state '%s', using 'default' (%s)",
--                 start_state, controld_globals.our_nodename);
-+                 start_state, node_name);
-     }
- }
- 
-@@ -335,7 +334,8 @@ do_cl_join_finalize_respond(long long action,
- 
-             first_join = FALSE;
-             if (start_state) {
--                set_join_state(start_state);
-+                set_join_state(start_state, controld_globals.our_nodename,
-+                               controld_globals.our_uuid);
-             }
-         }
- 
-diff --git a/daemons/controld/pacemaker-controld.h b/daemons/controld/pacemaker-controld.h
-index 1484a00..d8c2ddd 100644
---- a/daemons/controld/pacemaker-controld.h
-+++ b/daemons/controld/pacemaker-controld.h
-@@ -36,4 +36,7 @@ void controld_remove_voter(const char *uname);
- void controld_election_fini(void);
- void controld_stop_current_election_timeout(void);
- 
-+void set_join_state(const char *start_state, const char *node_name,
-+                    const char *node_uuid);
-+
- #endif
--- 
-2.31.1
-
-From 63d069adb344bba2c982013226f87dfd95afaff3 Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 13:38:03 -0400
-Subject: [PATCH 5/7] Refactor: controller: set_join_state needs to take a
- remote parameter.
-
-Without this parameter, we won't know what to pass to as node_type to
-cib__update_node_attr.  And without that, that function will not know to
-update a remote node - it'll try to update a regular node by the same
-name, which either doesn't exist or is not what we were hoping would
-happen.
-
-Ref T138
----
- daemons/controld/controld_join_client.c | 11 +++++++----
- daemons/controld/pacemaker-controld.h   |  2 +-
- 2 files changed, 8 insertions(+), 5 deletions(-)
-
-diff --git a/daemons/controld/controld_join_client.c b/daemons/controld/controld_join_client.c
-index 07e2a27..799d1b4 100644
---- a/daemons/controld/controld_join_client.c
-+++ b/daemons/controld/controld_join_client.c
-@@ -196,7 +196,8 @@ join_query_callback(xmlNode * msg, int call_id, int rc, xmlNode * output, void *
- }
- 
- void
--set_join_state(const char *start_state, const char *node_name, const char *node_uuid)
-+set_join_state(const char *start_state, const char *node_name, const char *node_uuid,
-+               bool remote)
- {
-     if (pcmk__str_eq(start_state, "standby", pcmk__str_casei)) {
-         crm_notice("Forcing node %s to join in %s state per configured "
-@@ -204,7 +205,8 @@ set_join_state(const char *start_state, const char *node_name, const char *node_
-         cib__update_node_attr(controld_globals.logger_out,
-                               controld_globals.cib_conn, cib_sync_call,
-                               XML_CIB_TAG_NODES, node_uuid,
--                              NULL, NULL, NULL, "standby", "on", NULL, NULL);
-+                              NULL, NULL, NULL, "standby", "on", NULL,
-+                              remote ? "remote" : NULL);
- 
-     } else if (pcmk__str_eq(start_state, "online", pcmk__str_casei)) {
-         crm_notice("Forcing node %s to join in %s state per configured "
-@@ -212,7 +214,8 @@ set_join_state(const char *start_state, const char *node_name, const char *node_
-         cib__update_node_attr(controld_globals.logger_out,
-                               controld_globals.cib_conn, cib_sync_call,
-                               XML_CIB_TAG_NODES, node_uuid,
--                              NULL, NULL, NULL, "standby", "off", NULL, NULL);
-+                              NULL, NULL, NULL, "standby", "off", NULL,
-+                              remote ? "remote" : NULL);
- 
-     } else if (pcmk__str_eq(start_state, "default", pcmk__str_casei)) {
-         crm_debug("Not forcing a starting state on node %s", node_name);
-@@ -335,7 +338,7 @@ do_cl_join_finalize_respond(long long action,
-             first_join = FALSE;
-             if (start_state) {
-                 set_join_state(start_state, controld_globals.our_nodename,
--                               controld_globals.our_uuid);
-+                               controld_globals.our_uuid, false);
-             }
-         }
- 
-diff --git a/daemons/controld/pacemaker-controld.h b/daemons/controld/pacemaker-controld.h
-index d8c2ddd..2334cce 100644
---- a/daemons/controld/pacemaker-controld.h
-+++ b/daemons/controld/pacemaker-controld.h
-@@ -37,6 +37,6 @@ void controld_election_fini(void);
- void controld_stop_current_election_timeout(void);
- 
- void set_join_state(const char *start_state, const char *node_name,
--                    const char *node_uuid);
-+                    const char *node_uuid, bool remote);
- 
- #endif
--- 
-2.31.1
-
-From 67274787898355065315f8c06d62458e2c2b0afe Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 10:09:02 -0400
-Subject: [PATCH 6/7] Feature: controller: When a remote node starts, apply any
- start state.
-
-If we were given a start state in the handshake XML, that is now stored
-in the remote node cache's private data.  Extract it and set the state
-on the node with set_node_state.
-
-Fixes T183
----
- daemons/controld/controld_remote_ra.c | 15 +++++++++++++++
- 1 file changed, 15 insertions(+)
-
-diff --git a/daemons/controld/controld_remote_ra.c b/daemons/controld/controld_remote_ra.c
-index f24b755..8ab1e46 100644
---- a/daemons/controld/controld_remote_ra.c
-+++ b/daemons/controld/controld_remote_ra.c
-@@ -280,6 +280,7 @@ remote_node_up(const char *node_name)
-     int call_opt;
-     xmlNode *update, *state;
-     crm_node_t *node;
-+    lrm_state_t *connection_rsc = NULL;
- 
-     CRM_CHECK(node_name != NULL, return);
-     crm_info("Announcing Pacemaker Remote node %s", node_name);
-@@ -301,6 +302,20 @@ remote_node_up(const char *node_name)
-     purge_remote_node_attrs(call_opt, node);
-     pcmk__update_peer_state(__func__, node, CRM_NODE_MEMBER, 0);
- 
-+    /* Apply any start state that we were given from the environment on the
-+     * remote node.
-+     */
-+    connection_rsc = lrm_state_find(node->uname);
-+
-+    if (connection_rsc != NULL) {
-+        lrmd_t *lrm = connection_rsc->conn;
-+        const char *start_state = lrmd__node_start_state(lrm);
-+
-+        if (start_state) {
-+            set_join_state(start_state, node->uname, node->uuid, true);
-+        }
-+    }
-+
-     /* pacemaker_remote nodes don't participate in the membership layer,
-      * so cluster nodes don't automatically get notified when they come and go.
-      * We send a cluster message to the DC, and update the CIB node state entry,
--- 
-2.31.1
-
-From 91cdda7056c9b9254a0d7e7a016b30f788e3e3ff Mon Sep 17 00:00:00 2001
-From: Chris Lumens <clumens@redhat.com>
-Date: Tue, 2 May 2023 10:16:30 -0400
-Subject: [PATCH 7/7] Doc: sysconfig: Remote nodes now respect start state.
-
-Ref T183
----
- etc/sysconfig/pacemaker.in | 3 +--
- 1 file changed, 1 insertion(+), 2 deletions(-)
-
-diff --git a/etc/sysconfig/pacemaker.in b/etc/sysconfig/pacemaker.in
-index 3b03ad6..041da71 100644
---- a/etc/sysconfig/pacemaker.in
-+++ b/etc/sysconfig/pacemaker.in
-@@ -144,8 +144,7 @@
- # By default, the local host will join the cluster in an online or standby
- # state when Pacemaker first starts depending on whether it was previously put
- # into standby mode. If this variable is set to "standby" or "online", it will
--# force the local host to join in the specified state. This has no effect on
--# Pacemaker Remote nodes.
-+# force the local host to join in the specified state.
- #
- # Default: PCMK_node_start_state="default"
- 
--- 
-2.31.1
-

SOURCES/002-remote-regression.patch

@@ -0,0 +1,98 @@
+From d8e08729ad5e3dc62f774172f992210902fc0ed4 Mon Sep 17 00:00:00 2001
+From: Ken Gaillot <kgaillot@redhat.com>
+Date: Mon, 23 Jan 2023 14:25:56 -0600
+Subject: [PATCH] High: executor: fix regression in remote node shutdown
+
+This reverts the essential part of d61494347, which was based on misdiagnosing
+a remote node shutdown issue. Initially, it was thought that a "TLS server
+session ended" log just after a remote node requested shutdown indicated that
+the proxy connection coincidentally dropped at that moment. It actually is the
+routine stopping of accepting new proxy connections, and existing when that
+happens makes the remote node exit immediately without waiting for the
+all-clear from the cluster.
+
+Fixes T361
+---
+ daemons/execd/pacemaker-execd.c | 19 +------------------
+ daemons/execd/pacemaker-execd.h |  3 +--
+ daemons/execd/remoted_tls.c     |  6 +-----
+ 3 files changed, 3 insertions(+), 25 deletions(-)
+
+diff --git a/daemons/execd/pacemaker-execd.c b/daemons/execd/pacemaker-execd.c
+index db12674f13..491808974a 100644
+--- a/daemons/execd/pacemaker-execd.c
++++ b/daemons/execd/pacemaker-execd.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2012-2022 the Pacemaker project contributors
++ * Copyright 2012-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -305,23 +305,6 @@ lrmd_exit(gpointer data)
+     return FALSE;
+ }
+ 
+-/*!
+- * \internal
+- * \brief Clean up and exit if shutdown has started
+- *
+- * \return Doesn't return
+- */
+-void
+-execd_exit_if_shutting_down(void)
+-{
+-#ifdef PCMK__COMPILE_REMOTE
+-    if (shutting_down) {
+-        crm_warn("exit because TLS connection was closed and 'shutting_down' set");
+-        lrmd_exit(NULL);
+-    }
+-#endif
+-}
+-
+ /*!
+  * \internal
+  * \brief Request cluster shutdown if appropriate, otherwise exit immediately
+diff --git a/daemons/execd/pacemaker-execd.h b/daemons/execd/pacemaker-execd.h
+index 6646ae29e3..f78e8dcdde 100644
+--- a/daemons/execd/pacemaker-execd.h
++++ b/daemons/execd/pacemaker-execd.h
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2012-2022 the Pacemaker project contributors
++ * Copyright 2012-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -105,6 +105,5 @@ void remoted_spawn_pidone(int argc, char **argv, char **envp);
+ int process_lrmd_alert_exec(pcmk__client_t *client, uint32_t id,
+                             xmlNode *request);
+ void lrmd_drain_alerts(GMainLoop *mloop);
+-void execd_exit_if_shutting_down(void);
+ 
+ #endif // PACEMAKER_EXECD__H
+diff --git a/daemons/execd/remoted_tls.c b/daemons/execd/remoted_tls.c
+index 6f4b2d0062..c65e3f394d 100644
+--- a/daemons/execd/remoted_tls.c
++++ b/daemons/execd/remoted_tls.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2012-2022 the Pacemaker project contributors
++ * Copyright 2012-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -250,10 +250,6 @@ static void
+ tls_server_dropped(gpointer user_data)
+ {
+     crm_notice("TLS server session ended");
+-    /* If we are in the process of shutting down, then we should actually exit.
+-     * bz#1804259
+-     */
+-    execd_exit_if_shutting_down();
+     return;
+ }
+ 
+-- 
+2.31.1
+

SOURCES/004-clone-rsc-display.patch

@@ -1,35 +0,0 @@
-From 770d417e28dc9527fec8b8a00caaba8825995454 Mon Sep 17 00:00:00 2001
-From: Grace Chin <gchin@redhat.com>
-Date: Wed, 19 Jul 2023 10:25:55 -0400
-Subject: [PATCH] Fix: tools: Fix a bug in clone resource description display
-
-Previously, descriptions of resources running on multiple
-nodes were displayed despite --full not being used (with pcs
-status) or --show-detail not being used (with crm_mon).
-
-For example, clone resources running on multiple nodes were
-affected.
-
-Now, --full and --show-detail must be used in order for resource
-descriptions to be displayed, regardless of the number of nodes
-the resource is run on.
-
-see bz: 2106642
----
- lib/pengine/pe_output.c | 3 +--
- 1 file changed, 1 insertion(+), 2 deletions(-)
-
-diff --git a/lib/pengine/pe_output.c b/lib/pengine/pe_output.c
-index e0b43d997a..d1c9f6e226 100644
---- a/lib/pengine/pe_output.c
-+++ b/lib/pengine/pe_output.c
-@@ -20,8 +20,7 @@ pe__resource_description(const pe_resource_t *rsc, uint32_t show_opts)
- {
-     const char * desc = NULL;
-     // User-supplied description
--    if (pcmk_any_flags_set(show_opts, pcmk_show_rsc_only|pcmk_show_description)
--        || pcmk__list_of_multiple(rsc->running_on)) {
-+    if (pcmk_any_flags_set(show_opts, pcmk_show_rsc_only|pcmk_show_description)) {
-         desc = crm_element_value(rsc->xml, XML_ATTR_DESC);
-     }
-     return desc;

SOURCES/004-g_source_remove.patch

@@ -0,0 +1,107 @@
+From 45617b727e280cac384a28ae3d96145e066e6197 Mon Sep 17 00:00:00 2001
+From: Reid Wahl <nrwahl@protonmail.com>
+Date: Fri, 3 Feb 2023 12:08:57 -0800
+Subject: [PATCH 01/02] Fix: fencer: Prevent double g_source_remove of op_timer_one
+
+QE observed a rarely reproducible core dump in the fencer during
+Pacemaker shutdown, in which we try to g_source_remove() an op timer
+that's already been removed.
+
+free_stonith_remote_op_list()
+-> g_hash_table_destroy()
+-> g_hash_table_remove_all_nodes()
+-> clear_remote_op_timers()
+-> g_source_remove()
+-> crm_glib_handler()
+-> "Source ID 190 was not found when attempting to remove it"
+
+The likely cause is that request_peer_fencing() doesn't set
+op->op_timer_one to 0 after calling g_source_remove() on it, so if that
+op is still in the stonith_remote_op_list at shutdown with the same
+timer, clear_remote_op_timers() tries to remove the source for
+op_timer_one again.
+
+There are only five locations that call g_source_remove() on a
+remote_fencing_op_t timer.
+* Three of them are in clear_remote_op_timers(), which first 0-checks
+  the timer and then sets it to 0 after g_source_remove().
+* One is in remote_op_query_timeout(), which does the same.
+* The last is the one we fix here in request_peer_fencing().
+
+I don't know all the conditions of QE's test scenario at this point.
+What I do know:
+* have-watchdog=true
+* stonith-watchdog-timeout=10
+* no explicit topology
+* fence agent script is missing for the configured fence device
+* requested fencing of one node
+* cluster shutdown
+
+Fixes RHBZ2166967
+
+Signed-off-by: Reid Wahl <nrwahl@protonmail.com>
+---
+ daemons/fenced/fenced_remote.c | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index d61b5bd..b7426ff 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -1825,6 +1825,7 @@ request_peer_fencing(remote_fencing_op_t *op, peer_device_info_t *peer)
+         op->state = st_exec;
+         if (op->op_timer_one) {
+             g_source_remove(op->op_timer_one);
++            op->op_timer_one = 0;
+         }
+ 
+         if (!((stonith_watchdog_timeout_ms > 0)
+-- 
+2.31.1
+
+From 0291db4750322ec7f01ae6a4a2a30abca9d8e19e Mon Sep 17 00:00:00 2001
+From: Reid Wahl <nrwahl@protonmail.com>
+Date: Wed, 15 Feb 2023 22:30:27 -0800
+Subject: [PATCH 02/02] Fix: fencer: Avoid double source remove of op_timer_total
+
+remote_op_timeout() returns G_SOURCE_REMOVE, which tells GLib to remove
+the source from the main loop after returning. Currently this function
+is used as the callback only when creating op->op_timer_total.
+
+If we don't set op->op_timer_total to 0 before returning from
+remote_op_timeout(), then we can get an assertion and core dump from
+GLib when the op's timers are being cleared (either during op
+finalization or during fencer shutdown). This is because
+clear_remote_op_timers() sees that op->op_timer_total != 0 and tries to
+remove the source, but the source has already been removed.
+
+Note that we're already (correctly) zeroing op->op_timer_one and
+op->query_timeout as appropriate in their respective callback functions.
+
+Fortunately, GLib doesn't care whether the source has already been
+removed before we return G_SOURCE_REMOVE from a callback. So it's safe
+to call finalize_op() (which removes all the op's timer sources) from
+within a callback.
+
+Fixes RHBZ#2166967
+
+Signed-off-by: Reid Wahl <nrwahl@protonmail.com>
+---
+ daemons/fenced/fenced_remote.c | 2 ++
+ 1 file changed, 2 insertions(+)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index b7426ff88..adea3d7d8 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -718,6 +718,8 @@ remote_op_timeout(gpointer userdata)
+ {
+     remote_fencing_op_t *op = userdata;
+ 
++    op->op_timer_total = 0;
++
+     if (op->state == st_done) {
+         crm_debug("Action '%s' targeting %s for client %s already completed "
+                   CRM_XS " id=%.8s",
+-- 
+2.39.0

SOURCES/005-query-null.patch

@@ -0,0 +1,151 @@
+From 0d15568a538349ac41028db6b506d13dd23e8732 Mon Sep 17 00:00:00 2001
+From: Chris Lumens <clumens@redhat.com>
+Date: Tue, 14 Feb 2023 14:00:37 -0500
+Subject: [PATCH] High: libcrmcommon: Fix handling node=NULL in
+ pcmk__attrd_api_query.
+
+According to the header file, if node is NULL, pcmk__attrd_api_query
+should query the value of the given attribute on all cluster nodes.
+This is also what the server expects and how attrd_updater is supposed
+to work.
+
+However, pcmk__attrd_api_query has no way of letting callers decide
+whether they want to query all nodes or whether they want to use the
+local node.  We were passing NULL for the node name, which it took to
+mean it should look up the local node name.  This calls
+pcmk__node_attr_target, which probes the local cluster name and returns
+that to pcmk__attrd_api_query.  If it returns non-NULL, that value will
+then be put into the XML IPC call which means the server will only
+return the value for that node.
+
+In testing this was usually fine.  However, in pratice, the methods
+pcmk__node_attr_target uses to figure out the local cluster node name
+involves checking the OCF_RESKEY_CRM_meta_on_node environment variable
+among others.
+
+This variable was never set in testing, but can be set in the real
+world.  This leads to circumstances where the user did "attrd_updater -QA"
+expecting to get the values on all nodes, but instead only got the value
+on the local cluster node.
+
+In pacemaker-2.1.4 and prior, pcmk__node_attr_target was simply never
+called if the node was NULL but was called otherwise.
+
+The fix is to modify pcmk__attrd_api_query to take an option for
+querying all nodes.  If that's present, we'll query all nodes.  If it's
+not present, we'll look at the given node name - NULL means look it up,
+anything else means just that node.
+
+Regression in 2.1.5 introduced by eb20a65577
+---
+ include/crm/common/attrd_internal.h     |  6 +++++-
+ include/crm/common/ipc_attrd_internal.h |  7 +++++--
+ lib/common/ipc_attrd.c                  | 12 ++++++++----
+ tools/attrd_updater.c                   |  5 +++--
+ 4 files changed, 21 insertions(+), 9 deletions(-)
+
+diff --git a/include/crm/common/attrd_internal.h b/include/crm/common/attrd_internal.h
+index 389be48..7337c38 100644
+--- a/include/crm/common/attrd_internal.h
++++ b/include/crm/common/attrd_internal.h
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2004-2022 the Pacemaker project contributors
++ * Copyright 2004-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -25,6 +25,10 @@ enum pcmk__node_attr_opts {
+     pcmk__node_attr_perm           = (1 << 5),
+     pcmk__node_attr_sync_local     = (1 << 6),
+     pcmk__node_attr_sync_cluster   = (1 << 7),
++    // pcmk__node_attr_utilization is 8, but that has not been backported.
++    // I'm leaving the gap here in case we backport that in the future and
++    // also to avoid problems on mixed-version clusters.
++    pcmk__node_attr_query_all      = (1 << 9),
+ };
+ 
+ #define pcmk__set_node_attr_flags(node_attr_flags, flags_to_set) do {   \
+diff --git a/include/crm/common/ipc_attrd_internal.h b/include/crm/common/ipc_attrd_internal.h
+index 2c6713f..b1b7584 100644
+--- a/include/crm/common/ipc_attrd_internal.h
++++ b/include/crm/common/ipc_attrd_internal.h
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2022 the Pacemaker project contributors
++ * Copyright 2022-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -110,10 +110,13 @@ int pcmk__attrd_api_purge(pcmk_ipc_api_t *api, const char *node);
+  *
+  * \param[in,out] api           Connection to pacemaker-attrd
+  * \param[in]     node          Look up the attribute for this node
+- *                              (or NULL for all nodes)
++ *                              (or NULL for the local node)
+  * \param[in]     name          Attribute name
+  * \param[in]     options       Bitmask of pcmk__node_attr_opts
+  *
++ * \note Passing pcmk__node_attr_query_all will cause the function to query
++ *       the value of \p name on all nodes, regardless of the value of \p node.
++ *
+  * \return Standard Pacemaker return code
+  */
+ int pcmk__attrd_api_query(pcmk_ipc_api_t *api, const char *node, const char *name,
+diff --git a/lib/common/ipc_attrd.c b/lib/common/ipc_attrd.c
+index 4606509..dece49b 100644
+--- a/lib/common/ipc_attrd.c
++++ b/lib/common/ipc_attrd.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2011-2022 the Pacemaker project contributors
++ * Copyright 2011-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -332,10 +332,14 @@ pcmk__attrd_api_query(pcmk_ipc_api_t *api, const char *node, const char *name,
+         return EINVAL;
+     }
+ 
+-    target = pcmk__node_attr_target(node);
++    if (pcmk_is_set(options, pcmk__node_attr_query_all)) {
++        node = NULL;
++    } else {
++        target = pcmk__node_attr_target(node);
+ 
+-    if (target != NULL) {
+-        node = target;
++        if (target != NULL) {
++            node = target;
++        }
+     }
+ 
+     request = create_attrd_op(NULL);
+diff --git a/tools/attrd_updater.c b/tools/attrd_updater.c
+index 3cd766d..cbd341d 100644
+--- a/tools/attrd_updater.c
++++ b/tools/attrd_updater.c
+@@ -376,6 +376,7 @@ attrd_event_cb(pcmk_ipc_api_t *attrd_api, enum pcmk_ipc_event event_type,
+ static int
+ send_attrd_query(pcmk__output_t *out, const char *attr_name, const char *attr_node, gboolean query_all)
+ {
++    uint32_t options = pcmk__node_attr_none;
+     pcmk_ipc_api_t *attrd_api = NULL;
+     int rc = pcmk_rc_ok;
+ 
+@@ -400,10 +401,10 @@ send_attrd_query(pcmk__output_t *out, const char *attr_name, const char *attr_no
+ 
+     /* Decide which node(s) to query */
+     if (query_all == TRUE) {
+-        attr_node = NULL;
++        options |= pcmk__node_attr_query_all;
+     }
+ 
+-    rc = pcmk__attrd_api_query(attrd_api, attr_node, attr_name, 0);
++    rc = pcmk__attrd_api_query(attrd_api, attr_node, attr_name, options);
+ 
+     if (rc != pcmk_rc_ok) {
+         g_set_error(&error, PCMK__RC_ERROR, rc, "Could not query value of %s: %s (%d)",
+-- 
+2.31.1
+

SOURCES/006-watchdog-fencing-topology.patch

@@ -0,0 +1,142 @@
+From 17cc49e1564b0ae55cc8212d14c5c055f88040da Mon Sep 17 00:00:00 2001
+From: Klaus Wenninger <klaus.wenninger@aon.at>
+Date: Tue, 14 Feb 2023 15:35:37 +0100
+Subject: [PATCH] Fix: watchdog-fencing: terminate dangling timer before
+ watchdog-waiting
+
+---
+ daemons/fenced/fenced_remote.c | 6 +++++-
+ 1 file changed, 5 insertions(+), 1 deletion(-)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index 5c3fe25e3..aab185adb 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2009-2022 the Pacemaker project contributors
++ * Copyright 2009-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -1702,6 +1702,10 @@ check_watchdog_fencing_and_wait(remote_fencing_op_t * op)
+                    "client %s " CRM_XS " id=%.8s",
+                    (stonith_watchdog_timeout_ms / 1000),
+                    op->target, op->action, op->client_name, op->id);
++
++        if (op->op_timer_one) {
++            g_source_remove(op->op_timer_one);
++        }
+         op->op_timer_one = g_timeout_add(stonith_watchdog_timeout_ms,
+                                          remote_op_watchdog_done, op);
+         return TRUE;
+-- 
+2.39.0
+
+From f2cc2a4277124230903a18713e50604a8f1842cd Mon Sep 17 00:00:00 2001
+From: Klaus Wenninger <klaus.wenninger@aon.at>
+Date: Wed, 1 Mar 2023 15:00:15 +0100
+Subject: [PATCH] Refactor: watchdog-fencing: convenience function
+ pcmk__is_fencing_action
+
+for consistency and add comment making clear why this block exits
+with new timer set in any case
+---
+ daemons/fenced/fenced_remote.c | 5 ++++-
+ 1 file changed, 4 insertions(+), 1 deletion(-)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index aab185adb..e0f8de057 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -1834,7 +1834,7 @@ request_peer_fencing(remote_fencing_op_t *op, peer_device_info_t *peer)
+         if (!((stonith_watchdog_timeout_ms > 0)
+               && (pcmk__str_eq(device, STONITH_WATCHDOG_ID, pcmk__str_none)
+                   || (pcmk__str_eq(peer->host, op->target, pcmk__str_casei)
+-                      && !pcmk__str_eq(op->action, "on", pcmk__str_none)))
++                      && pcmk__is_fencing_action(op->action)))
+               && check_watchdog_fencing_and_wait(op))) {
+ 
+             /* Some thoughts about self-fencing cases reaching this point:
+@@ -1854,6 +1854,9 @@ request_peer_fencing(remote_fencing_op_t *op, peer_device_info_t *peer)
+                  Otherwise the selection of stonith-watchdog-timeout at
+                  least is questionable.
+              */
++
++            /* coming here we're not waiting for watchdog timeout -
++               thus engage timer with timout evaluated before */
+             op->op_timer_one = g_timeout_add((1000 * timeout_one), remote_op_timeout_one, op);
+         }
+ 
+-- 
+2.39.0
+
+From c4eb45a986f8865fc5e69350fd5b9f4b056d9d69 Mon Sep 17 00:00:00 2001
+From: Klaus Wenninger <klaus.wenninger@aon.at>
+Date: Tue, 14 Feb 2023 11:57:17 +0100
+Subject: [PATCH] Fix: watchdog-fencing: correctly derive timeout with topology
+
+up to now the timeout for watchdog-fencing was just added to
+the overall timeout if the node to be fenced was visible and
+reported back to the query.
+---
+ daemons/fenced/fenced_remote.c | 28 +++++++++++++++++++++++++---
+ 1 file changed, 25 insertions(+), 3 deletions(-)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index e0f8de057..3b7ab05e9 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -969,8 +969,9 @@ advance_topology_level(remote_fencing_op_t *op, bool empty_ok)
+         return pcmk_rc_ok;
+     }
+ 
+-    crm_info("All fencing options targeting %s for client %s@%s failed "
++    crm_info("All %sfencing options targeting %s for client %s@%s failed "
+              CRM_XS " id=%.8s",
++             (stonith_watchdog_timeout_ms > 0)?"non-watchdog ":"",
+              op->target, op->client_name, op->originator, op->id);
+     return ENODEV;
+ }
+@@ -1434,8 +1435,17 @@ stonith_choose_peer(remote_fencing_op_t * op)
+              && pcmk_is_set(op->call_options, st_opt_topology)
+              && (advance_topology_level(op, false) == pcmk_rc_ok));
+ 
+-    crm_notice("Couldn't find anyone to fence (%s) %s using %s",
+-               op->action, op->target, (device? device : "any device"));
++    if ((stonith_watchdog_timeout_ms > 0)
++        && pcmk__is_fencing_action(op->action)
++        && pcmk__str_eq(device, STONITH_WATCHDOG_ID, pcmk__str_none)
++        && node_does_watchdog_fencing(op->target)) {
++        crm_info("Couldn't contact watchdog-fencing target-node (%s)",
++                 op->target);
++        /* check_watchdog_fencing_and_wait will log additional info */
++    } else {
++        crm_notice("Couldn't find anyone to fence (%s) %s using %s",
++                   op->action, op->target, (device? device : "any device"));
++    }
+     return NULL;
+ }
+ 
+@@ -1531,6 +1541,18 @@ get_op_total_timeout(const remote_fencing_op_t *op,
+                 continue;
+             }
+             for (device_list = tp->levels[i]; device_list; device_list = device_list->next) {
++                /* in case of watchdog-device we add the timeout to the budget
++                   regardless of if we got a reply or not
++                 */
++                if ((stonith_watchdog_timeout_ms > 0)
++                    && pcmk__is_fencing_action(op->action)
++                    && pcmk__str_eq(device_list->data, STONITH_WATCHDOG_ID,
++                                    pcmk__str_none)
++                    && node_does_watchdog_fencing(op->target)) {
++                    total_timeout += stonith_watchdog_timeout_ms / 1000;
++                    continue;
++                }
++
+                 for (iter = op->query_results; iter != NULL; iter = iter->next) {
+                     const peer_device_info_t *peer = iter->data;
+ 
+-- 
+2.39.0
+

SOURCES/007-attrd-dampen.patch

@@ -12,7 +12,7 @@ diff --git a/tools/attrd_updater.c b/tools/attrd_updater.c
 index b615a3575..4688b9ff6 100644
 --- a/tools/attrd_updater.c
 +++ b/tools/attrd_updater.c
-@@ -501,7 +501,7 @@ send_attrd_update(char command, const char *attr_node, const char *attr_name,
+@@ -445,7 +445,7 @@ send_attrd_update(char command, const char *attr_node, const char *attr_name,
  
          case 'U':
              rc = pcmk__attrd_api_update(NULL, attr_node, attr_name, attr_value,

SOURCES/007-glib-assertions.patch

@@ -1,163 +0,0 @@
-From 63f4bd4d5a324e6eb279340a42c7c36c8902ada7 Mon Sep 17 00:00:00 2001
-From: Ken Gaillot <kgaillot@redhat.com>
-Date: Wed, 2 Aug 2023 15:55:26 -0500
-Subject: [PATCH 1/4] Fix: controller: don't try to execute agent action at
- shutdown
-
-Normally, agent execution is not possible at shutdown. However, when metadata
-is needed for some action, the agent can be called asynchronously, and when the
-metadata action returns, the original action is performed. If the metadata is
-initiated before shutdown, but completes after shutdown has begun, do not try
-to attempt the original action, so we avoid unnecessary error logs.
----
- daemons/controld/controld_execd.c | 4 +++-
- 1 file changed, 3 insertions(+), 1 deletion(-)
-
-diff --git a/daemons/controld/controld_execd.c b/daemons/controld/controld_execd.c
-index 530e4346c8..a90e8d833e 100644
---- a/daemons/controld/controld_execd.c
-+++ b/daemons/controld/controld_execd.c
-@@ -1400,7 +1400,9 @@ metadata_complete(int pid, const pcmk__action_result_t *result, void *user_data)
-         md = controld_cache_metadata(lrm_state->metadata_cache, data->rsc,
-                                      result->action_stdout);
-     }
--    do_lrm_rsc_op(lrm_state, data->rsc, data->input_xml, md);
-+    if (!pcmk_is_set(controld_globals.fsa_input_register, R_HA_DISCONNECTED)) {
-+        do_lrm_rsc_op(lrm_state, data->rsc, data->input_xml, md);
-+    }
-     free_metadata_cb_data(data);
- }
- 
-
-From 247d9534f36f690c1474e36cedaadb3934022a05 Mon Sep 17 00:00:00 2001
-From: Ken Gaillot <kgaillot@redhat.com>
-Date: Wed, 2 Aug 2023 16:16:31 -0500
-Subject: [PATCH 2/4] Refactor: controller: de-functionize lrm_state_destroy()
-
-It was a one-liner called once
----
- daemons/controld/controld_execd_state.c | 8 +-------
- daemons/controld/controld_lrm.h         | 5 -----
- 2 files changed, 1 insertion(+), 12 deletions(-)
-
-diff --git a/daemons/controld/controld_execd_state.c b/daemons/controld/controld_execd_state.c
-index 8c68bfca08..4a87a9b332 100644
---- a/daemons/controld/controld_execd_state.c
-+++ b/daemons/controld/controld_execd_state.c
-@@ -132,12 +132,6 @@ lrm_state_create(const char *node_name)
-     return state;
- }
- 
--void
--lrm_state_destroy(const char *node_name)
--{
--    g_hash_table_remove(lrm_state_table, node_name);
--}
--
- static gboolean
- remote_proxy_remove_by_node(gpointer key, gpointer value, gpointer user_data)
- {
-@@ -799,7 +793,7 @@ lrm_state_unregister_rsc(lrm_state_t * lrm_state,
-     }
- 
-     if (is_remote_lrmd_ra(NULL, NULL, rsc_id)) {
--        lrm_state_destroy(rsc_id);
-+        g_hash_table_remove(lrm_state_table, rsc_id);
-         return pcmk_ok;
-     }
- 
-diff --git a/daemons/controld/controld_lrm.h b/daemons/controld/controld_lrm.h
-index 25f3db3316..c3113e49c3 100644
---- a/daemons/controld/controld_lrm.h
-+++ b/daemons/controld/controld_lrm.h
-@@ -108,11 +108,6 @@ gboolean lrm_state_init_local(void);
-  */
- void lrm_state_destroy_all(void);
- 
--/*!
-- * \brief Destroy executor connection by node name
-- */
--void lrm_state_destroy(const char *node_name);
--
- /*!
-  * \brief Find lrm_state data by node name
-  */
-
-From 1b915f1ce38756431f7faa142565e3e07aade194 Mon Sep 17 00:00:00 2001
-From: Ken Gaillot <kgaillot@redhat.com>
-Date: Wed, 2 Aug 2023 15:58:09 -0500
-Subject: [PATCH 3/4] Low: controller: guard lrm_state_table usage with NULL
- check
-
-It is NULL while draining the mainloop during the shutdown sequence.
----
- daemons/controld/controld_execd_state.c | 7 ++++++-
- 1 file changed, 6 insertions(+), 1 deletion(-)
-
-diff --git a/daemons/controld/controld_execd_state.c b/daemons/controld/controld_execd_state.c
-index 4a87a9b332..b90cc5e635 100644
---- a/daemons/controld/controld_execd_state.c
-+++ b/daemons/controld/controld_execd_state.c
-@@ -301,7 +301,7 @@ lrm_state_destroy_all(void)
- lrm_state_t *
- lrm_state_find(const char *node_name)
- {
--    if (!node_name) {
-+    if ((node_name == NULL) || (lrm_state_table == NULL)) {
-         return NULL;
-     }
-     return g_hash_table_lookup(lrm_state_table, node_name);
-@@ -312,6 +312,8 @@ lrm_state_find_or_create(const char *node_name)
- {
-     lrm_state_t *lrm_state;
- 
-+    CRM_CHECK(lrm_state_table != NULL, return NULL);
-+
-     lrm_state = g_hash_table_lookup(lrm_state_table, node_name);
-     if (!lrm_state) {
-         lrm_state = lrm_state_create(node_name);
-@@ -323,6 +325,9 @@ lrm_state_find_or_create(const char *node_name)
- GList *
- lrm_state_get_list(void)
- {
-+    if (lrm_state_table == NULL) {
-+        return NULL;
-+    }
-     return g_hash_table_get_values(lrm_state_table);
- }
- 
-
-From 78581213ed3bf4183b0ec1f391b720d5d91f3f68 Mon Sep 17 00:00:00 2001
-From: Ken Gaillot <kgaillot@redhat.com>
-Date: Wed, 2 Aug 2023 15:48:36 -0500
-Subject: [PATCH 4/4] Log: controller: improve messages for resource history
- updates
-
----
- daemons/controld/controld_cib.c | 11 +++++++++--
- 1 file changed, 9 insertions(+), 2 deletions(-)
-
-diff --git a/daemons/controld/controld_cib.c b/daemons/controld/controld_cib.c
-index 22ac42486f..c9dde0b748 100644
---- a/daemons/controld/controld_cib.c
-+++ b/daemons/controld/controld_cib.c
-@@ -861,10 +861,17 @@ cib_rsc_callback(xmlNode * msg, int call_id, int rc, xmlNode * output, void *use
-         case pcmk_ok:
-         case -pcmk_err_diff_failed:
-         case -pcmk_err_diff_resync:
--            crm_trace("Resource update %d complete: rc=%d", call_id, rc);
-+            crm_trace("Resource history update completed (call=%d rc=%d)",
-+                      call_id, rc);
-             break;
-         default:
--            crm_warn("Resource update %d failed: (rc=%d) %s", call_id, rc, pcmk_strerror(rc));
-+            if (call_id > 0) {
-+                crm_warn("Resource history update %d failed: %s "
-+                         CRM_XS " rc=%d", call_id, pcmk_strerror(rc), rc);
-+            } else {
-+                crm_warn("Resource history update failed: %s " CRM_XS " rc=%d",
-+                         pcmk_strerror(rc), rc);
-+            }
-     }
- 
-     if (call_id == pending_rsc_update) {

SOURCES/008-controller-reply.patch

@@ -26,7 +26,7 @@ diff --git a/lib/common/ipc_controld.c b/lib/common/ipc_controld.c
 index 3c3a98964..405fd0518 100644
 --- a/lib/common/ipc_controld.c
 +++ b/lib/common/ipc_controld.c
-@@ -177,18 +177,16 @@ set_nodes_data(pcmk_controld_api_reply_t *data, xmlNode *msg_data)
+@@ -143,18 +143,16 @@ set_nodes_data(pcmk_controld_api_reply_t *data, xmlNode *msg_data)
  static bool
  reply_expected(pcmk_ipc_api_t *api, xmlNode *request)
  {
@@ -55,7 +55,7 @@ index 3c3a98964..405fd0518 100644
  }
  
  static bool
-@@ -202,22 +200,12 @@ dispatch(pcmk_ipc_api_t *api, xmlNode *reply)
+@@ -168,22 +166,12 @@ dispatch(pcmk_ipc_api_t *api, xmlNode *reply)
          pcmk_controld_reply_unknown, NULL, NULL,
      };
  
@@ -83,7 +83,7 @@ index 3c3a98964..405fd0518 100644
      }
  
      if (private->replies_expected > 0) {
-@@ -344,18 +332,15 @@ static int
+@@ -310,18 +298,15 @@ static int
  send_controller_request(pcmk_ipc_api_t *api, xmlNode *request,
                          bool reply_is_expected)
  {

SOURCES/009-glib-assertions.patch

@@ -0,0 +1,127 @@
+From 2587f9fabea3a7ef01eb7752d4e2ef082823934e Mon Sep 17 00:00:00 2001
+From: eabdullin <ed.abdullin.1@gmail.com>
+Date: Wed, 13 Sep 2023 14:15:46 +0300
+Subject: [PATCH] - Fix: controller: don't try to execute agent action at
+ shutdown Normally, agent execution is not possible at shutdown. However, when
+ metadata is needed for some action, the agent can be called asynchronously,
+ and when the metadata action returns, the original action is performed. If
+ the metadata is initiated before shutdown, but completes after shutdown has
+ begun, do not try to attempt the original action, so we avoid unnecessary
+ error logs. - Refactor: controller: de-functionize lrm_state_destroy() It was
+ a one-liner called once - Log: controller: improve messages for resource
+ history updates - Low: controller: guard lrm_state_table usage with NULLcheck
+ It is NULL while draining the mainloop during the shutdown sequence.
+
+---
+ daemons/controld/controld_execd.c       | 15 ++++++++++++---
+ daemons/controld/controld_execd_state.c | 15 +++++++--------
+ daemons/controld/controld_lrm.h         |  5 -----
+ 3 files changed, 19 insertions(+), 16 deletions(-)
+
+diff --git a/daemons/controld/controld_execd.c b/daemons/controld/controld_execd.c
+index afead92..e7a91ab 100644
+--- a/daemons/controld/controld_execd.c
++++ b/daemons/controld/controld_execd.c
+@@ -1728,7 +1728,9 @@ metadata_complete(int pid, const pcmk__action_result_t *result, void *user_data)
+         md = controld_cache_metadata(lrm_state->metadata_cache, data->rsc,
+                                      result->action_stdout);
+     }
+-    do_lrm_rsc_op(lrm_state, data->rsc, data->input_xml, md);
++    if (!pcmk_is_set(fsa_input_register, R_HA_DISCONNECTED)) {
++        do_lrm_rsc_op(lrm_state, data->rsc, data->input_xml, md);
++    }
+     free_metadata_cb_data(data);
+ }
+ 
+@@ -2406,10 +2408,17 @@ cib_rsc_callback(xmlNode * msg, int call_id, int rc, xmlNode * output, void *use
+         case pcmk_ok:
+         case -pcmk_err_diff_failed:
+         case -pcmk_err_diff_resync:
+-            crm_trace("Resource update %d complete: rc=%d", call_id, rc);
++            crm_trace("Resource history update completed (call=%d rc=%d)",
++                      call_id, rc);
+             break;
+         default:
+-            crm_warn("Resource update %d failed: (rc=%d) %s", call_id, rc, pcmk_strerror(rc));
++            if (call_id > 0) {
++                crm_warn("Resource history update %d failed: %s "
++                         CRM_XS " rc=%d", call_id, pcmk_strerror(rc), rc);
++            } else {
++                crm_warn("Resource history update failed: %s " CRM_XS " rc=%d",
++                         pcmk_strerror(rc), rc);
++            }
+     }
+ 
+     if (call_id == last_resource_update) {
+diff --git a/daemons/controld/controld_execd_state.c b/daemons/controld/controld_execd_state.c
+index adba2e5..3994b6d 100644
+--- a/daemons/controld/controld_execd_state.c
++++ b/daemons/controld/controld_execd_state.c
+@@ -131,12 +131,6 @@ lrm_state_create(const char *node_name)
+     return state;
+ }
+ 
+-void
+-lrm_state_destroy(const char *node_name)
+-{
+-    g_hash_table_remove(lrm_state_table, node_name);
+-}
+-
+ static gboolean
+ remote_proxy_remove_by_node(gpointer key, gpointer value, gpointer user_data)
+ {
+@@ -252,7 +246,7 @@ lrm_state_destroy_all(void)
+ lrm_state_t *
+ lrm_state_find(const char *node_name)
+ {
+-    if (!node_name) {
++    if ((node_name == NULL) || (lrm_state_table == NULL)) {
+         return NULL;
+     }
+     return g_hash_table_lookup(lrm_state_table, node_name);
+@@ -263,6 +257,8 @@ lrm_state_find_or_create(const char *node_name)
+ {
+     lrm_state_t *lrm_state;
+ 
++    CRM_CHECK(lrm_state_table != NULL, return NULL);
++
+     lrm_state = g_hash_table_lookup(lrm_state_table, node_name);
+     if (!lrm_state) {
+         lrm_state = lrm_state_create(node_name);
+@@ -274,6 +270,9 @@ lrm_state_find_or_create(const char *node_name)
+ GList *
+ lrm_state_get_list(void)
+ {
++    if (lrm_state_table == NULL) {
++        return NULL;
++    }
+     return g_hash_table_get_values(lrm_state_table);
+ }
+ 
+@@ -764,7 +763,7 @@ lrm_state_unregister_rsc(lrm_state_t * lrm_state,
+     }
+ 
+     if (is_remote_lrmd_ra(NULL, NULL, rsc_id)) {
+-        lrm_state_destroy(rsc_id);
++        g_hash_table_remove(lrm_state_table, rsc_id);
+         return pcmk_ok;
+     }
+ 
+diff --git a/daemons/controld/controld_lrm.h b/daemons/controld/controld_lrm.h
+index 983c288..11ff1bc 100644
+--- a/daemons/controld/controld_lrm.h
++++ b/daemons/controld/controld_lrm.h
+@@ -113,11 +113,6 @@ void lrm_state_destroy_all(void);
+  */
+ lrm_state_t *lrm_state_create(const char *node_name);
+ 
+-/*!
+- * \brief Destroy executor connection by node name
+- */
+-void lrm_state_destroy(const char *node_name);
+-
+ /*!
+  * \brief Find lrm_state data by node name
+  */
+-- 
+

SOURCES/010-revert-58400e27.patch

@@ -1,62 +0,0 @@
-From 2e81e0db9a716c486805e0760f78be65ca79eeae Mon Sep 17 00:00:00 2001
-From: Ken Gaillot <kgaillot@redhat.com>
-Date: Tue, 17 Oct 2023 15:28:27 -0500
-Subject: [PATCH] Fix: attrd: avoid regression by reverting 58400e27
-
-Fixes T714
----
- daemons/attrd/attrd_cib.c       |  5 -----
- daemons/attrd/attrd_elections.c | 10 +---------
- daemons/attrd/pacemaker-attrd.h |  1 -
- 3 files changed, 1 insertion(+), 15 deletions(-)
-
-diff --git a/daemons/attrd/attrd_cib.c b/daemons/attrd/attrd_cib.c
-index 2de37a7cb6..9ce2872715 100644
---- a/daemons/attrd/attrd_cib.c
-+++ b/daemons/attrd/attrd_cib.c
-@@ -641,11 +641,6 @@ attrd_write_attributes(uint32_t options)
-               pcmk_is_set(options, attrd_write_all)? "all" : "changed");
-     g_hash_table_iter_init(&iter, attributes);
-     while (g_hash_table_iter_next(&iter, NULL, (gpointer *) & a)) {
--        if (pcmk_is_set(options, attrd_write_skip_shutdown)
--            && pcmk__str_eq(a->id, XML_CIB_ATTR_SHUTDOWN, pcmk__str_none)) {
--            continue;
--        }
--
-         if (!pcmk_is_set(options, attrd_write_all) && a->unknown_peer_uuids) {
-             // Try writing this attribute again, in case peer ID was learned
-             a->changed = true;
-diff --git a/daemons/attrd/attrd_elections.c b/daemons/attrd/attrd_elections.c
-index a95cd44cbd..62310ed1d8 100644
---- a/daemons/attrd/attrd_elections.c
-+++ b/daemons/attrd/attrd_elections.c
-@@ -35,16 +35,8 @@ attrd_election_cb(gpointer user_data)
- 
-     /* After winning an election, update the CIB with the values of all
-      * attributes as the winner knows them.
--     *
--     * However, do not write out any "shutdown" attributes. A node that is
--     * shutting down will have all its transient attributes removed from the CIB
--     * when its controller exits, and from the attribute manager's memory (on
--     * remaining nodes) when its attribute manager exits; if an election is won
--     * between when those two things happen, we don't want to write the shutdown
--     * attribute back out, which would cause the node to immediately shut down
--     * the next time it rejoins.
-      */
--    attrd_write_attributes(attrd_write_all|attrd_write_skip_shutdown);
-+    attrd_write_attributes(attrd_write_all);
-     return G_SOURCE_REMOVE;
- }
- 
-diff --git a/daemons/attrd/pacemaker-attrd.h b/daemons/attrd/pacemaker-attrd.h
-index e3c369b5bc..a95bb54367 100644
---- a/daemons/attrd/pacemaker-attrd.h
-+++ b/daemons/attrd/pacemaker-attrd.h
-@@ -181,7 +181,6 @@ enum attrd_write_options {
-     attrd_write_changed         = 0,
-     attrd_write_all             = (1 << 0),
-     attrd_write_no_delay        = (1 << 1),
--    attrd_write_skip_shutdown   = (1 << 2),
- };
- 
- void attrd_write_attribute(attribute_t *a, bool ignore_delay);

SOURCES/011-revert-f5263c94.patch

@@ -1,34 +0,0 @@
-From 14b87a38786ae5b4dc12fc1581e5d39a274fced2 Mon Sep 17 00:00:00 2001
-From: Ken Gaillot <kgaillot@redhat.com>
-Date: Mon, 30 Oct 2023 12:21:24 -0500
-Subject: [PATCH] Fix: attrd: revert faulty T138 fix
-
-f5263c9401 created a timing issue where a node could get a shutdown attribute,
-the original writer leaves the cluster before writing it out, then the
-shutting-down node wins the writer election. In that case, it would skip the
-write-out and the scheduler would never shut it down.
-
-Reopens T138
----
- daemons/attrd/attrd_elections.c | 8 --------
- 1 file changed, 8 deletions(-)
-
-diff --git a/daemons/attrd/attrd_elections.c b/daemons/attrd/attrd_elections.c
-index 62310ed1d8..82fbe8affc 100644
---- a/daemons/attrd/attrd_elections.c
-+++ b/daemons/attrd/attrd_elections.c
-@@ -22,14 +22,6 @@
- {
-     attrd_declare_winner();
- 
--    if (attrd_requesting_shutdown() || attrd_shutting_down()) {
--        /* This node is shutting down or about to, meaning its attributes will
--         * be removed (and may have already been removed from the CIB by a
--         * controller). Don't sync or write its attributes in this case.
--         */
--        return G_SOURCE_REMOVE;
--    }
--
-     /* Update the peers after an election */
-     attrd_peer_sync(NULL, NULL);
-