The branch, v4-6-test has been updated
via f10ce74 ctdb-recovery: Do not run local ip verification when in
recovery
via 9f25dff ctdb-recovery: Get recmode unconditionally in the main_loop
via 59ac9bf ctdb-recovery: Finish processing for recovery mode ACTIVE
first
via 7ee7e65 ctdb-recovery: Simplify logging of recovery mode setting
via 89ee737 ctdb-recovery: Setting up of recmode should be idempotent
via a227893 ctdb-recovery: Assign banning credits if database fails to
freeze
via 6e11262 ctdb-scripts: Don't send empty argument string to logger
from 9670a0d Bug 15852. There are valid paths where
conn->lsa_pipe_tcp->transport is NULL. Protect against this.
https://git.samba.org/?p=samba.git;a=shortlog;h=v4-6-test
- Log -----------------------------------------------------------------
commit f10ce74c333d1943e58456b600eac6eefb1767bb
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 16:15:47 2017 +1000
ctdb-recovery: Do not run local ip verification when in recovery
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
If we drop public IPs because CTDB is in recovery for too long, then
avoid spamming logs "Trigger takeoverrun" every second.
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
(cherry picked from commit ea46699b27ef8d4ac7b5dd07035465cb3df09ea4)
Autobuild-User(v4-6-test): Karolin Seeger <[email protected]>
Autobuild-Date(v4-6-test): Mon Jun 26 16:51:56 CEST 2017 on sn-devel-144
commit 9f25dffb4f2370888a20ce4d7d988bed99c1e676
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 17:45:20 2017 +1000
ctdb-recovery: Get recmode unconditionally in the main_loop
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
This can be used later in the main_loop to avoid the local ip check.
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
(cherry picked from commit 2fd2ccd4c8617cfa7374d7a5ee3d1cc61c4fa4ad)
commit 59ac9bfe6006a4e8596771af2c7ff3bf4512f4f4
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 14:09:32 2017 +1000
ctdb-recovery: Finish processing for recovery mode ACTIVE first
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
This simplifies the code and avoids complicated conditions.
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
(cherry picked from commit f8200153b21f5b19c9a1d57be3e05e739d9fafcd)
commit 7ee7e65f5212adefc5ea9351e30c6487b83dd0f1
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 14:52:32 2017 +1000
ctdb-recovery: Simplify logging of recovery mode setting
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
(cherry picked from commit d74dadd7f26a9e8c48ba92468d7d0c4a7aa5a8e5)
commit 89ee737cd41d49855f881d519f6736cfeaba9178
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 14:49:02 2017 +1000
ctdb-recovery: Setting up of recmode should be idempotent
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12857
If the recovery mode is already set to the expected value, there is
nothing to do.
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
(cherry picked from commit f2771fcbf438e8b06321752c7203f01bbe33b573)
commit a227893188bbf5c68dcf923d42ccd20ae366009a
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 14:00:13 2017 +1000
ctdb-recovery: Assign banning credits if database fails to freeze
https://bugzilla.samba.org/show_bug.cgi?id=12857
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
(cherry picked from commit c9d9f56bffe1e19665dba8e0cf899399d3d9fb72)
commit 6e11262d60bfbc8e053de10a188beed52dd6d90a
Author: Amitay Isaacs <[email protected]>
Date: Thu Jun 22 14:34:36 2017 +1000
ctdb-scripts: Don't send empty argument string to logger
BUG: https://bugzilla.samba.org/show_bug.cgi?id=12856
This stops logger reading from stdin.
Signed-off-by: Amitay Isaacs <[email protected]>
Reviewed-by: Martin Schwenke <[email protected]>
Autobuild-User(master): Martin Schwenke <[email protected]>
Autobuild-Date(master): Sat Jun 24 14:37:48 CEST 2017 on sn-devel-144
(cherry picked from commit 2fa069e870c0e6f1daa2a08be5aacac7b1063a5e)
-----------------------------------------------------------------------
Summary of changes:
ctdb/config/functions | 2 +-
ctdb/server/ctdb_recover.c | 28 ++++++++++++++--------------
ctdb/server/ctdb_recoverd.c | 19 ++++++++++++-------
ctdb/server/ctdb_recovery_helper.c | 1 +
4 files changed, 28 insertions(+), 22 deletions(-)
Changeset truncated at 500 lines:
diff --git a/ctdb/config/functions b/ctdb/config/functions
index 7e37bbb..3826324 100755
--- a/ctdb/config/functions
+++ b/ctdb/config/functions
@@ -150,7 +150,7 @@ script_log ()
*)
# Handle all syslog:* variants here too. There's no tool to do
# the lossy things, so just use logger.
- logger -t "ctdbd: ${_tag}" "$*"
+ logger -t "ctdbd: ${_tag}" "$@"
;;
esac
}
diff --git a/ctdb/server/ctdb_recover.c b/ctdb/server/ctdb_recover.c
index 6bed61c..813a1ad 100644
--- a/ctdb/server/ctdb_recover.c
+++ b/ctdb/server/ctdb_recover.c
@@ -856,26 +856,24 @@ int32_t ctdb_control_set_recmode(struct ctdb_context
*ctdb,
struct set_recmode_state *state;
struct ctdb_cluster_mutex_handle *h;
+ if (recmode == ctdb->recovery_mode) {
+ D_INFO("Recovery mode already set to %s\n",
+ recmode == CTDB_RECOVERY_NORMAL ? "NORMAL" : "ACTIVE");
+ return 0;
+ }
+
+ D_NOTICE("Recovery mode set to %s\n",
+ recmode == CTDB_RECOVERY_NORMAL ? "NORMAL" : "ACTIVE");
+
/* if we enter recovery but stay in recovery for too long
we will eventually drop all our ip addresses
*/
- if (recmode == CTDB_RECOVERY_NORMAL) {
- talloc_free(ctdb->release_ips_ctx);
- ctdb->release_ips_ctx = NULL;
- } else {
+ if (recmode == CTDB_RECOVERY_ACTIVE) {
if (ctdb_deferred_drop_all_ips(ctdb) != 0) {
- DEBUG(DEBUG_ERR,("Failed to set up deferred drop all
ips\n"));
+ D_ERR("Failed to set up deferred drop all ips\n");
}
- }
- if (recmode != ctdb->recovery_mode) {
- DEBUG(DEBUG_NOTICE,(__location__ " Recovery mode set to %s\n",
- recmode==CTDB_RECOVERY_NORMAL?"NORMAL":"ACTIVE"));
- }
-
- if (recmode != CTDB_RECOVERY_NORMAL ||
- ctdb->recovery_mode != CTDB_RECOVERY_ACTIVE) {
- ctdb->recovery_mode = recmode;
+ ctdb->recovery_mode = CTDB_RECOVERY_ACTIVE;
return 0;
}
@@ -884,6 +882,8 @@ int32_t ctdb_control_set_recmode(struct ctdb_context *ctdb,
* Therefore, what follows is special handling when setting
* recovery mode back to normal */
+ TALLOC_FREE(ctdb->release_ips_ctx);
+
for (ctdb_db = ctdb->db_list; ctdb_db != NULL; ctdb_db = ctdb_db->next)
{
if (ctdb_db->generation != ctdb->vnn_map->generation) {
DEBUG(DEBUG_ERR,
diff --git a/ctdb/server/ctdb_recoverd.c b/ctdb/server/ctdb_recoverd.c
index 9ea0f61..d9cc4a2 100644
--- a/ctdb/server/ctdb_recoverd.c
+++ b/ctdb/server/ctdb_recoverd.c
@@ -2608,6 +2608,13 @@ static void main_loop(struct ctdb_context *ctdb, struct
ctdb_recoverd *rec,
return;
}
+ ret = ctdb_ctrl_getrecmode(ctdb, mem_ctx, CONTROL_TIMEOUT(),
+ CTDB_CURRENT_NODE, &ctdb->recovery_mode);
+ if (ret != 0) {
+ D_ERR("Failed to read recmode from local node\n");
+ return;
+ }
+
/* if the local daemon is STOPPED or BANNED, we verify that the
databases are
also frozen and that the recmode is set to active.
*/
@@ -2620,10 +2627,6 @@ static void main_loop(struct ctdb_context *ctdb, struct
ctdb_recoverd *rec,
*/
rec->priority_time = timeval_current();
- ret = ctdb_ctrl_getrecmode(ctdb, mem_ctx, CONTROL_TIMEOUT(),
CTDB_CURRENT_NODE, &ctdb->recovery_mode);
- if (ret != 0) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to read recmode
from local node\n"));
- }
if (ctdb->recovery_mode == CTDB_RECOVERY_NORMAL) {
DEBUG(DEBUG_ERR,("Node is stopped or banned but
recovery mode is not active. Activate recovery mode and lock databases\n"));
@@ -2667,9 +2670,11 @@ static void main_loop(struct ctdb_context *ctdb, struct
ctdb_recoverd *rec,
return;
}
- /* Check if an IP takeover run is needed and trigger one if
- * necessary */
- verify_local_ip_allocation(ctdb, rec, pnn, nodemap);
+ if (ctdb->recovery_mode == CTDB_RECOVERY_NORMAL) {
+ /* Check if an IP takeover run is needed and trigger one if
+ * necessary */
+ verify_local_ip_allocation(ctdb, rec, pnn, nodemap);
+ }
/* if we are not the recmaster then we do not need to check
if recovery is needed
diff --git a/ctdb/server/ctdb_recovery_helper.c
b/ctdb/server/ctdb_recovery_helper.c
index 0222aa0..474b900 100644
--- a/ctdb/server/ctdb_recovery_helper.c
+++ b/ctdb/server/ctdb_recovery_helper.c
@@ -1627,6 +1627,7 @@ static void recover_db_freeze_done(struct tevent_req
*subreq)
if (ret2 != 0) {
LOG("control FREEZE_DB failed for db %s on node %u,"
" ret=%d\n", state->db_name, pnn, ret2);
+ state->ban_credits[pnn] += 1;
} else {
LOG("control FREEZE_DB failed for db %s, ret=%d\n",
state->db_name, ret);
--
Samba Shared Repository
