[SCM] CTDB repository - branch master updated - ctdb-1.0.114-15-g2a79d97
Ronnie Sahlberg
sahlberg at samba.org
Wed Apr 7 22:40:07 MDT 2010
The branch, master has been updated
via 2a79d976f975ec1ecbd06c210cbb5e6a93068cec (commit)
via 8825a8e3936e4ddfc1482d07a63a620249f01fe7 (commit)
via f3bf2ab61f8dbbc806ec23a68a87aaedd458e712 (commit)
from 685be0a7e830464e01dccc744362040a75bc96b5 (commit)
http://gitweb.samba.org/?p=sahlberg/ctdb.git;a=shortlog;h=master
- Log -----------------------------------------------------------------
commit 2a79d976f975ec1ecbd06c210cbb5e6a93068cec
Merge: 685be0a7e830464e01dccc744362040a75bc96b5 8825a8e3936e4ddfc1482d07a63a620249f01fe7
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Apr 8 14:30:01 2010 +1000
Merge root at 10.1.1.27:/shared/ctdb/ctdb-git
commit 8825a8e3936e4ddfc1482d07a63a620249f01fe7
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Apr 8 14:28:52 2010 +1000
Fix a compiler warning
commit f3bf2ab61f8dbbc806ec23a68a87aaedd458e712
Author: Ronnie Sahlberg <ronniesahlberg at gmail.com>
Date: Thu Apr 8 14:07:57 2010 +1000
In the recovery daemon, keep track of which node we have assigned public ip
addresses and verify that the remote nodes have/keep a consistent view of
assigned addresses.
If a remote node has an inconsistent view of addresses visavi the recovery
master this will trigger a full ip reallocation.
-----------------------------------------------------------------------
Summary of changes:
client/ctdb_client.c | 2 +-
common/rb_tree.h | 12 +++++-----
include/ctdb_private.h | 9 ++++++-
server/ctdb_recoverd.c | 18 ++++++++++-----
server/ctdb_takeover.c | 57 ++++++++++++++++++++++++++++++++++++++++++------
5 files changed, 77 insertions(+), 21 deletions(-)
Changeset truncated at 500 lines:
diff --git a/client/ctdb_client.c b/client/ctdb_client.c
index 9e5f7fc..946335c 100644
--- a/client/ctdb_client.c
+++ b/client/ctdb_client.c
@@ -1906,7 +1906,7 @@ int ctdb_dumpdb_record(struct ctdb_context *ctdb, TDB_DATA key, TDB_DATA data, v
fprintf(f, "dmaster: %u\n", h->dmaster);
fprintf(f, "rsn: %llu\n", (unsigned long long)h->rsn);
- fprintf(f, "data(%u) = \"", (unsigned)data.dsize - sizeof(*h));
+ fprintf(f, "data(%u) = \"", (unsigned)(data.dsize - sizeof(*h)));
for (i=sizeof(*h);i<data.dsize;i++) {
if (ISASCII(data.dptr[i])) {
fprintf(f, "%c", data.dptr[i]);
diff --git a/common/rb_tree.h b/common/rb_tree.h
index cb7cba3..eef0bc5 100644
--- a/common/rb_tree.h
+++ b/common/rb_tree.h
@@ -21,17 +21,17 @@
#define TRBT_RED 0x00
#define TRBT_BLACK 0x01
-typedef struct _trbt_node_t {
- struct _trbt_tree_t *tree;
- struct _trbt_node_t *parent;
- struct _trbt_node_t *left;
- struct _trbt_node_t *right;
+typedef struct trbt_node {
+ struct trbt_tree *tree;
+ struct trbt_node *parent;
+ struct trbt_node *left;
+ struct trbt_node *right;
uint32_t rb_color;
uint32_t key32;
void *data;
} trbt_node_t;
-typedef struct _trbt_tree_t {
+typedef struct trbt_tree {
trbt_node_t *root;
/* automatically free the tree when the last node has been deleted */
#define TRBT_AUTOFREE 0x00000001
diff --git a/include/ctdb_private.h b/include/ctdb_private.h
index 953552a..888b626 100644
--- a/include/ctdb_private.h
+++ b/include/ctdb_private.h
@@ -442,7 +442,7 @@ struct ctdb_context {
struct ctdb_call_state *pending_calls;
struct ctdb_client_ip *client_ip_list;
bool do_checkpublicip;
- struct _trbt_tree_t *server_ids;
+ struct trbt_tree *server_ids;
const char *event_script_dir;
const char *notification_script;
const char *default_public_interface;
@@ -469,6 +469,9 @@ struct ctdb_context {
/* mapping from pid to ctdb_client * */
struct ctdb_client_pid_list *client_pids;
+
+ /* used in the recovery daemon to remember the ip allocation */
+ struct trbt_tree *ip_tree;
};
struct ctdb_db_context {
@@ -1633,4 +1636,8 @@ int ctdb_recheck_persistent_health(struct ctdb_context *ctdb);
void ctdb_run_notification_script(struct ctdb_context *ctdb, const char *event);
void ctdb_fault_setup(void);
+
+int verify_remote_ip_allocation(struct ctdb_context *ctdb,
+ struct ctdb_all_public_ips *ips);
+
#endif
diff --git a/server/ctdb_recoverd.c b/server/ctdb_recoverd.c
index f5aa77b..5f34711 100644
--- a/server/ctdb_recoverd.c
+++ b/server/ctdb_recoverd.c
@@ -1228,6 +1228,7 @@ static void reload_nodes_file(struct ctdb_context *ctdb)
}
static int ctdb_reload_remote_public_ips(struct ctdb_context *ctdb,
+ struct ctdb_recoverd *rec,
struct ctdb_node_map *nodemap,
uint32_t *culprit)
{
@@ -1274,6 +1275,11 @@ static int ctdb_reload_remote_public_ips(struct ctdb_context *ctdb,
return -1;
}
+ if (verify_remote_ip_allocation(ctdb, ctdb->nodes[j]->known_public_ips)) {
+ DEBUG(DEBUG_ERR,("Node %d has inconsistent public ip allocation and needs update.\n", ctdb->nodes[j]->pnn));
+ rec->need_takeover_run = true;
+ }
+
/* grab a new shiny list of public ips from the node */
ret = ctdb_ctrl_get_public_ips_flags(ctdb,
CONTROL_TIMEOUT(),
@@ -1568,7 +1574,7 @@ static int do_recovery(struct ctdb_recoverd *rec,
/*
tell nodes to takeover their public IPs
*/
- ret = ctdb_reload_remote_public_ips(ctdb, nodemap, &culprit);
+ ret = ctdb_reload_remote_public_ips(ctdb, rec, nodemap, &culprit);
if (ret != 0) {
DEBUG(DEBUG_ERR,("Failed to read public ips from remote node %d\n",
culprit));
@@ -1961,7 +1967,7 @@ static void process_ipreallocate_requests(struct ctdb_context *ctdb, struct ctdb
/* update the list of public ips that a node can handle for
all connected nodes
*/
- ret = ctdb_reload_remote_public_ips(ctdb, rec->nodemap, &culprit);
+ ret = ctdb_reload_remote_public_ips(ctdb, rec, rec->nodemap, &culprit);
if (ret != 0) {
DEBUG(DEBUG_ERR,("Failed to read public ips from remote node %d\n",
culprit));
@@ -2414,9 +2420,9 @@ static enum monitor_result verify_recmaster(struct ctdb_recoverd *rec, struct ct
}
-/* called to check that the allocation of public ip addresses is ok.
+/* called to check that the local allocation of public ip addresses is ok.
*/
-static int verify_ip_allocation(struct ctdb_context *ctdb, struct ctdb_recoverd *rec, uint32_t pnn)
+static int verify_local_ip_allocation(struct ctdb_context *ctdb, struct ctdb_recoverd *rec, uint32_t pnn)
{
TALLOC_CTX *mem_ctx = talloc_new(NULL);
struct ctdb_control_get_ifaces *ifaces = NULL;
@@ -3088,7 +3094,7 @@ again:
*/
if (ctdb->do_checkpublicip) {
if (rec->ip_check_disable_ctx == NULL) {
- if (verify_ip_allocation(ctdb, rec, pnn) != 0) {
+ if (verify_local_ip_allocation(ctdb, rec, pnn) != 0) {
DEBUG(DEBUG_ERR, (__location__ " Public IPs were inconsistent.\n"));
}
}
@@ -3353,7 +3359,7 @@ again:
/* update the list of public ips that a node can handle for
all connected nodes
*/
- ret = ctdb_reload_remote_public_ips(ctdb, nodemap, &culprit);
+ ret = ctdb_reload_remote_public_ips(ctdb, rec, nodemap, &culprit);
if (ret != 0) {
DEBUG(DEBUG_ERR,("Failed to read public ips from remote node %d\n",
culprit));
diff --git a/server/ctdb_takeover.c b/server/ctdb_takeover.c
index 166ca5a..5433172 100644
--- a/server/ctdb_takeover.c
+++ b/server/ctdb_takeover.c
@@ -1159,14 +1159,17 @@ void getips_count_callback(void *param, void *data)
}
struct ctdb_public_ip_list *
-create_merged_ip_list(struct ctdb_context *ctdb, TALLOC_CTX *tmp_ctx)
+create_merged_ip_list(struct ctdb_context *ctdb)
{
int i, j;
struct ctdb_public_ip_list *ip_list;
struct ctdb_all_public_ips *public_ips;
- trbt_tree_t *ip_tree;
- ip_tree = trbt_create(tmp_ctx, 0);
+ if (ctdb->ip_tree != NULL) {
+ talloc_free(ctdb->ip_tree);
+ ctdb->ip_tree = NULL;
+ }
+ ctdb->ip_tree = trbt_create(ctdb, 0);
for (i=0;i<ctdb->num_nodes;i++) {
public_ips = ctdb->nodes[i]->known_public_ips;
@@ -1183,13 +1186,13 @@ create_merged_ip_list(struct ctdb_context *ctdb, TALLOC_CTX *tmp_ctx)
for (j=0;j<public_ips->num;j++) {
struct ctdb_public_ip_list *tmp_ip;
- tmp_ip = talloc_zero(tmp_ctx, struct ctdb_public_ip_list);
+ tmp_ip = talloc_zero(ctdb->ip_tree, struct ctdb_public_ip_list);
CTDB_NO_MEMORY_NULL(ctdb, tmp_ip);
tmp_ip->pnn = public_ips->ips[j].pnn;
tmp_ip->addr = public_ips->ips[j].addr;
tmp_ip->next = NULL;
- trbt_insertarray32_callback(ip_tree,
+ trbt_insertarray32_callback(ctdb->ip_tree,
IP_KEYLEN, ip_key(&public_ips->ips[j].addr),
add_ip_callback,
tmp_ip);
@@ -1197,7 +1200,7 @@ create_merged_ip_list(struct ctdb_context *ctdb, TALLOC_CTX *tmp_ctx)
}
ip_list = NULL;
- trbt_traversearray32(ip_tree, IP_KEYLEN, getips_count_callback, &ip_list);
+ trbt_traversearray32(ctdb->ip_tree, IP_KEYLEN, getips_count_callback, &ip_list);
return ip_list;
}
@@ -1247,8 +1250,10 @@ int ctdb_takeover_run(struct ctdb_context *ctdb, struct ctdb_node_map *nodemap)
a full list of all public addresses that exist in the cluster.
Walk over all node structures and create a merged list of
all public addresses that exist in the cluster.
+
+ keep the tree of ips around as ctdb->ip_tree
*/
- all_ips = create_merged_ip_list(ctdb, tmp_ctx);
+ all_ips = create_merged_ip_list(ctdb);
/* If we want deterministic ip allocations, i.e. that the ip addresses
will always be allocated the same way for a specific set of
@@ -2806,3 +2811,41 @@ int32_t ctdb_control_del_public_address(struct ctdb_context *ctdb, TDB_DATA inda
return -1;
}
+/* This function is called from the recovery daemon to verify that a remote
+ node has the expected ip allocation.
+ This is verified against ctdb->ip_tree
+*/
+int verify_remote_ip_allocation(struct ctdb_context *ctdb, struct ctdb_all_public_ips *ips)
+{
+ struct ctdb_public_ip_list *tmp_ip;
+ int i;
+
+ if (ctdb->ip_tree == NULL) {
+ /* dont know the expected allocation yet, assume remote node
+ is correct. */
+ return 0;
+ }
+
+ if (ips == NULL) {
+ return 0;
+ }
+
+ for (i=0; i<ips->num; i++) {
+ tmp_ip = trbt_lookuparray32(ctdb->ip_tree, IP_KEYLEN, ip_key(&ips->ips[i].addr));
+ if (tmp_ip == NULL) {
+ DEBUG(DEBUG_ERR,(__location__ " Could not find host for address %s, reassign ips\n", ctdb_addr_to_str(&ips->ips[i].addr)));
+ return -1;
+ }
+
+ if (tmp_ip->pnn == -1 || ips->ips[i].pnn == -1) {
+ continue;
+ }
+
+ if (tmp_ip->pnn != ips->ips[i].pnn) {
+ DEBUG(DEBUG_ERR,("Inconsistent ip allocation. Trigger reallocation.\n"));
+ return -1;
+ }
+ }
+
+ return 0;
+}
--
CTDB repository
More information about the samba-cvs
mailing list