#include "db_wrap.h"
#include "lib/tdb/include/tdb.h"
#include "lib/util/dlinklist.h"
-#include "lib/events/events.h"
+#include "lib/tevent/tevent.h"
#include "system/network.h"
#include "system/filesys.h"
#include "system/locale.h"
#include "../include/ctdb_private.h"
#include "lib/util/dlinklist.h"
+pid_t ctdbd_pid;
+
/*
allocate a packet for use in client<->daemon communication
*/
*/
int ctdb_call_local(struct ctdb_db_context *ctdb_db, struct ctdb_call *call,
struct ctdb_ltdb_header *header, TALLOC_CTX *mem_ctx,
- TDB_DATA *data, uint32_t caller)
+ TDB_DATA *data)
{
struct ctdb_call_info *c;
struct ctdb_registered_call *fn;
return -1;
}
- if (header->laccessor != caller) {
- header->lacount = 0;
- }
- header->laccessor = caller;
- header->lacount++;
-
- /* we need to force the record to be written out if this was a remote access,
- so that the lacount is updated */
- if (c->new_data == NULL && header->laccessor != ctdb->pnn) {
+ /* we need to force the record to be written out if this was a remote access */
+ if (c->new_data == NULL) {
c->new_data = &c->record_data;
}
ctdb->daemon.queue = ctdb_queue_setup(ctdb, ctdb, ctdb->daemon.sd,
CTDB_DS_ALIGNMENT,
- ctdb_client_read_cb, ctdb);
+ ctdb_client_read_cb, ctdb, "to-ctdbd");
return 0;
}
*(state->call) = *call;
state->ctdb_db = ctdb_db;
- ret = ctdb_call_local(ctdb_db, state->call, header, state, data, ctdb->pnn);
+ ret = ctdb_call_local(ctdb_db, state->call, header, state, data);
return state;
}
tell the daemon what messaging srvid we will use, and register the message
handler function in the client
*/
-int ctdb_set_message_handler(struct ctdb_context *ctdb, uint64_t srvid,
- ctdb_message_fn_t handler,
+int ctdb_client_set_message_handler(struct ctdb_context *ctdb, uint64_t srvid,
+ ctdb_msg_fn_t handler,
void *private_data)
{
/*
tell the daemon we no longer want a srvid
*/
-int ctdb_remove_message_handler(struct ctdb_context *ctdb, uint64_t srvid, void *private_data)
+int ctdb_client_remove_message_handler(struct ctdb_context *ctdb, uint64_t srvid, void *private_data)
{
int res;
int32_t status;
/*
send a message - from client context
*/
-int ctdb_send_message(struct ctdb_context *ctdb, uint32_t pnn,
+int ctdb_client_send_message(struct ctdb_context *ctdb, uint32_t pnn,
uint64_t srvid, TDB_DATA data)
{
struct ctdb_req_message *r;
{
struct ctdb_client_control_state *state = talloc_get_type(private_data, struct ctdb_client_control_state);
- DEBUG(DEBUG_ERR,("control timed out. reqid:%d opcode:%d dstnode:%d\n", state->reqid, state->c->opcode, state->c->hdr.destnode));
+ DEBUG(DEBUG_ERR,(__location__ " control timed out. reqid:%u opcode:%u "
+ "dstnode:%u\n", state->reqid, state->c->opcode,
+ state->c->hdr.destnode));
state->state = CTDB_CONTROL_TIMEOUT;
return 0;
}
+/*
+ get the health status of a db
+ */
+int ctdb_ctrl_getdbhealth(struct ctdb_context *ctdb,
+ struct timeval timeout,
+ uint32_t destnode,
+ uint32_t dbid, TALLOC_CTX *mem_ctx,
+ const char **reason)
+{
+ int ret;
+ int32_t res;
+ TDB_DATA data;
+
+ data.dptr = (uint8_t *)&dbid;
+ data.dsize = sizeof(dbid);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_DB_GET_HEALTH, 0, data,
+ mem_ctx, &data, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ return -1;
+ }
+
+ if (data.dsize == 0) {
+ (*reason) = NULL;
+ return 0;
+ }
+
+ (*reason) = talloc_strndup(mem_ctx, (const char *)data.dptr, data.dsize);
+ if ((*reason) == NULL) {
+ return -1;
+ }
+
+ talloc_free(data.dptr);
+
+ return 0;
+}
+
/*
create a database
*/
}
tdb_flags = persistent?TDB_DEFAULT:TDB_NOSYNC;
- if (!ctdb->do_setsched) {
+ if (ctdb->valgrinding) {
tdb_flags |= TDB_NOMMAP;
}
+ tdb_flags |= TDB_DISALLOW_NESTING;
ctdb_db->ltdb = tdb_wrap_open(ctdb, ctdb_db->db_path, 0, tdb_flags, O_RDWR, 0);
if (ctdb_db->ltdb == NULL) {
state.private_data = private_data;
state.fn = fn;
- ret = ctdb_set_message_handler(ctdb_db->ctdb, srvid, traverse_handler, &state);
+ ret = ctdb_client_set_message_handler(ctdb_db->ctdb, srvid, traverse_handler, &state);
if (ret != 0) {
DEBUG(DEBUG_ERR,("Failed to setup traverse handler\n"));
return -1;
data, NULL, NULL, &status, NULL, NULL);
if (ret != 0 || status != 0) {
DEBUG(DEBUG_ERR,("ctdb_traverse_all failed\n"));
- ctdb_remove_message_handler(ctdb_db->ctdb, srvid, &state);
+ ctdb_client_remove_message_handler(ctdb_db->ctdb, srvid, &state);
return -1;
}
event_loop_once(ctdb_db->ctdb->ev);
}
- ret = ctdb_remove_message_handler(ctdb_db->ctdb, srvid, &state);
+ ret = ctdb_client_remove_message_handler(ctdb_db->ctdb, srvid, &state);
if (ret != 0) {
DEBUG(DEBUG_ERR,("Failed to remove ctdb_traverse handler\n"));
return -1;
/*
called on each key during a catdb
*/
-static int dumpdb_fn(struct ctdb_context *ctdb, TDB_DATA key, TDB_DATA data, void *p)
+int ctdb_dumpdb_record(struct ctdb_context *ctdb, TDB_DATA key, TDB_DATA data, void *p)
{
int i;
FILE *f = (FILE *)p;
struct ctdb_ltdb_header *h = (struct ctdb_ltdb_header *)data.dptr;
- fprintf(f, "dmaster: %u\n", h->dmaster);
- fprintf(f, "rsn: %llu\n", (unsigned long long)h->rsn);
-
fprintf(f, "key(%u) = \"", (unsigned)key.dsize);
for (i=0;i<key.dsize;i++) {
if (ISASCII(key.dptr[i])) {
}
fprintf(f, "\"\n");
- fprintf(f, "data(%u) = \"", (unsigned)data.dsize);
+ 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)));
for (i=sizeof(*h);i<data.dsize;i++) {
if (ISASCII(data.dptr[i])) {
fprintf(f, "%c", data.dptr[i]);
}
fprintf(f, "\"\n");
+ fprintf(f, "\n");
+
return 0;
}
*/
int ctdb_dump_db(struct ctdb_db_context *ctdb_db, FILE *f)
{
- return ctdb_traverse(ctdb_db, dumpdb_fn, f);
+ return ctdb_traverse(ctdb_db, ctdb_dumpdb_record, f);
}
/*
async freeze send control
*/
struct ctdb_client_control_state *
-ctdb_ctrl_freeze_send(struct ctdb_context *ctdb, TALLOC_CTX *mem_ctx, struct timeval timeout, uint32_t destnode)
+ctdb_ctrl_freeze_send(struct ctdb_context *ctdb, TALLOC_CTX *mem_ctx, struct timeval timeout, uint32_t destnode, uint32_t priority)
{
- return ctdb_control_send(ctdb, destnode, 0,
+ return ctdb_control_send(ctdb, destnode, priority,
CTDB_CONTROL_FREEZE, 0, tdb_null,
mem_ctx, &timeout, NULL);
}
}
/*
- freeze a node
+ freeze databases of a certain priority
*/
-int ctdb_ctrl_freeze(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode)
+int ctdb_ctrl_freeze_priority(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, uint32_t priority)
{
TALLOC_CTX *tmp_ctx = talloc_new(ctdb);
struct ctdb_client_control_state *state;
int ret;
- state = ctdb_ctrl_freeze_send(ctdb, tmp_ctx, timeout, destnode);
+ state = ctdb_ctrl_freeze_send(ctdb, tmp_ctx, timeout, destnode, priority);
ret = ctdb_ctrl_freeze_recv(ctdb, tmp_ctx, state);
talloc_free(tmp_ctx);
return ret;
}
+/* Freeze all databases */
+int ctdb_ctrl_freeze(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode)
+{
+ int i;
+
+ for (i=1; i<=NUM_DB_PRIORITIES; i++) {
+ if (ctdb_ctrl_freeze_priority(ctdb, timeout, destnode, i) != 0) {
+ return -1;
+ }
+ }
+ return 0;
+}
+
/*
- thaw a node
+ thaw databases of a certain priority
*/
-int ctdb_ctrl_thaw(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode)
+int ctdb_ctrl_thaw_priority(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, uint32_t priority)
{
int ret;
int32_t res;
- ret = ctdb_control(ctdb, destnode, 0,
+ ret = ctdb_control(ctdb, destnode, priority,
CTDB_CONTROL_THAW, 0, tdb_null,
NULL, NULL, &res, &timeout, NULL);
if (ret != 0 || res != 0) {
return 0;
}
+/* thaw all databases */
+int ctdb_ctrl_thaw(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode)
+{
+ return ctdb_ctrl_thaw_priority(ctdb, timeout, destnode, 0);
+}
+
/*
get pnn of a node, or -1
*/
}
-int ctdb_ctrl_get_public_ips(struct ctdb_context *ctdb,
- struct timeval timeout, uint32_t destnode,
- TALLOC_CTX *mem_ctx, struct ctdb_all_public_ips **ips)
+int ctdb_ctrl_get_public_ips_flags(struct ctdb_context *ctdb,
+ struct timeval timeout, uint32_t destnode,
+ TALLOC_CTX *mem_ctx,
+ uint32_t flags,
+ struct ctdb_all_public_ips **ips)
{
int ret;
TDB_DATA outdata;
int32_t res;
ret = ctdb_control(ctdb, destnode, 0,
- CTDB_CONTROL_GET_PUBLIC_IPS, 0, tdb_null,
+ CTDB_CONTROL_GET_PUBLIC_IPS, flags, tdb_null,
mem_ctx, &outdata, &res, &timeout, NULL);
if (ret == 0 && res == -1) {
DEBUG(DEBUG_ERR,(__location__ " ctdb_control to get public ips failed, falling back to ipv4-only version\n"));
return 0;
}
+int ctdb_ctrl_get_public_ips(struct ctdb_context *ctdb,
+ struct timeval timeout, uint32_t destnode,
+ TALLOC_CTX *mem_ctx,
+ struct ctdb_all_public_ips **ips)
+{
+ return ctdb_ctrl_get_public_ips_flags(ctdb, timeout,
+ destnode, mem_ctx,
+ 0, ips);
+}
+
int ctdb_ctrl_get_public_ipsv4(struct ctdb_context *ctdb,
struct timeval timeout, uint32_t destnode,
TALLOC_CTX *mem_ctx, struct ctdb_all_public_ips **ips)
return 0;
}
+int ctdb_ctrl_get_public_ip_info(struct ctdb_context *ctdb,
+ struct timeval timeout, uint32_t destnode,
+ TALLOC_CTX *mem_ctx,
+ const ctdb_sock_addr *addr,
+ struct ctdb_control_public_ip_info **_info)
+{
+ int ret;
+ TDB_DATA indata;
+ TDB_DATA outdata;
+ int32_t res;
+ struct ctdb_control_public_ip_info *info;
+ uint32_t len;
+ uint32_t i;
+
+ indata.dptr = discard_const_p(uint8_t, addr);
+ indata.dsize = sizeof(*addr);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_GET_PUBLIC_IP_INFO, 0, indata,
+ mem_ctx, &outdata, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get public ip info "
+ "failed ret:%d res:%d\n",
+ ret, res));
+ return -1;
+ }
+
+ len = offsetof(struct ctdb_control_public_ip_info, ifaces);
+ if (len > outdata.dsize) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get public ip info "
+ "returned invalid data with size %u > %u\n",
+ (unsigned int)outdata.dsize,
+ (unsigned int)len));
+ dump_data(DEBUG_DEBUG, outdata.dptr, outdata.dsize);
+ return -1;
+ }
+
+ info = (struct ctdb_control_public_ip_info *)outdata.dptr;
+ len += info->num*sizeof(struct ctdb_control_iface_info);
+
+ if (len > outdata.dsize) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get public ip info "
+ "returned invalid data with size %u > %u\n",
+ (unsigned int)outdata.dsize,
+ (unsigned int)len));
+ dump_data(DEBUG_DEBUG, outdata.dptr, outdata.dsize);
+ return -1;
+ }
+
+ /* make sure we null terminate the returned strings */
+ for (i=0; i < info->num; i++) {
+ info->ifaces[i].name[CTDB_IFACE_SIZE] = '\0';
+ }
+
+ *_info = (struct ctdb_control_public_ip_info *)talloc_memdup(mem_ctx,
+ outdata.dptr,
+ outdata.dsize);
+ talloc_free(outdata.dptr);
+ if (*_info == NULL) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get public ip info "
+ "talloc_memdup size %u failed\n",
+ (unsigned int)outdata.dsize));
+ return -1;
+ }
+
+ return 0;
+}
+
+int ctdb_ctrl_get_ifaces(struct ctdb_context *ctdb,
+ struct timeval timeout, uint32_t destnode,
+ TALLOC_CTX *mem_ctx,
+ struct ctdb_control_get_ifaces **_ifaces)
+{
+ int ret;
+ TDB_DATA outdata;
+ int32_t res;
+ struct ctdb_control_get_ifaces *ifaces;
+ uint32_t len;
+ uint32_t i;
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_GET_IFACES, 0, tdb_null,
+ mem_ctx, &outdata, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get ifaces "
+ "failed ret:%d res:%d\n",
+ ret, res));
+ return -1;
+ }
+
+ len = offsetof(struct ctdb_control_get_ifaces, ifaces);
+ if (len > outdata.dsize) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get ifaces "
+ "returned invalid data with size %u > %u\n",
+ (unsigned int)outdata.dsize,
+ (unsigned int)len));
+ dump_data(DEBUG_DEBUG, outdata.dptr, outdata.dsize);
+ return -1;
+ }
+
+ ifaces = (struct ctdb_control_get_ifaces *)outdata.dptr;
+ len += ifaces->num*sizeof(struct ctdb_control_iface_info);
+
+ if (len > outdata.dsize) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get ifaces "
+ "returned invalid data with size %u > %u\n",
+ (unsigned int)outdata.dsize,
+ (unsigned int)len));
+ dump_data(DEBUG_DEBUG, outdata.dptr, outdata.dsize);
+ return -1;
+ }
+
+ /* make sure we null terminate the returned strings */
+ for (i=0; i < ifaces->num; i++) {
+ ifaces->ifaces[i].name[CTDB_IFACE_SIZE] = '\0';
+ }
+
+ *_ifaces = (struct ctdb_control_get_ifaces *)talloc_memdup(mem_ctx,
+ outdata.dptr,
+ outdata.dsize);
+ talloc_free(outdata.dptr);
+ if (*_ifaces == NULL) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for get ifaces "
+ "talloc_memdup size %u failed\n",
+ (unsigned int)outdata.dsize));
+ return -1;
+ }
+
+ return 0;
+}
+
+int ctdb_ctrl_set_iface_link(struct ctdb_context *ctdb,
+ struct timeval timeout, uint32_t destnode,
+ TALLOC_CTX *mem_ctx,
+ const struct ctdb_control_iface_info *info)
+{
+ int ret;
+ TDB_DATA indata;
+ int32_t res;
+
+ indata.dptr = discard_const_p(uint8_t, info);
+ indata.dsize = sizeof(*info);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_SET_IFACE_LINK_STATE, 0, indata,
+ mem_ctx, NULL, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for set iface link "
+ "failed ret:%d res:%d\n",
+ ret, res));
+ return -1;
+ }
+
+ return 0;
+}
+
/*
set/clear the permanent disabled bit on a remote node
*/
nodes = list_of_connected_nodes(ctdb, nodemap, tmp_ctx, true);
if (ctdb_client_async_control(ctdb, CTDB_CONTROL_MODIFY_FLAGS,
- nodes,
+ nodes, 0,
timeout, false, data,
NULL, NULL,
NULL) != 0) {
- DEBUG(DEBUG_ERR, (__location__ " ctdb_control to disable node failed\n"));
+ DEBUG(DEBUG_ERR, (__location__ " Unable to update nodeflags on remote nodes\n"));
talloc_free(tmp_ctx);
return -1;
}
ctdb->ev = ev;
ctdb->idr = idr_init(ctdb);
+ /* Wrap early to exercise code. */
+ ctdb->lastid = INT_MAX-200;
CTDB_NO_MEMORY_NULL(ctdb, ctdb->idr);
ret = ctdb_set_socketname(ctdb, CTDB_PATH);
return NULL;
}
+ ctdb->statistics.statistics_start_time = timeval_current();
+
return ctdb;
}
return 0;
}
+const char *ctdb_get_socketname(struct ctdb_context *ctdb)
+{
+ return ctdb->daemon.name;
+}
+
/*
return the pnn of this node
*/
int ctdb_client_async_control(struct ctdb_context *ctdb,
enum ctdb_controls opcode,
uint32_t *nodes,
+ uint64_t srvid,
struct timeval timeout,
bool dont_log_errors,
TDB_DATA data,
for (j=0; j<num_nodes; j++) {
uint32_t pnn = nodes[j];
- state = ctdb_control_send(ctdb, pnn, 0, opcode,
+ state = ctdb_control_send(ctdb, pnn, srvid, opcode,
0, data, async_data, &timeout, NULL);
if (state == NULL) {
DEBUG(DEBUG_ERR,(__location__ " Failed to call async control %u\n", (unsigned)opcode));
return nodes;
}
+uint32_t *list_of_active_nodes_except_pnn(struct ctdb_context *ctdb,
+ struct ctdb_node_map *node_map,
+ TALLOC_CTX *mem_ctx,
+ uint32_t pnn)
+{
+ int i, j, num_nodes;
+ uint32_t *nodes;
+
+ for (i=num_nodes=0;i<node_map->num;i++) {
+ if (node_map->nodes[i].flags & NODE_FLAGS_INACTIVE) {
+ continue;
+ }
+ if (node_map->nodes[i].pnn == pnn) {
+ continue;
+ }
+ num_nodes++;
+ }
+
+ nodes = talloc_array(mem_ctx, uint32_t, num_nodes);
+ CTDB_NO_MEMORY_FATAL(ctdb, nodes);
+
+ for (i=j=0;i<node_map->num;i++) {
+ if (node_map->nodes[i].flags & NODE_FLAGS_INACTIVE) {
+ continue;
+ }
+ if (node_map->nodes[i].pnn == pnn) {
+ continue;
+ }
+ nodes[j++] = node_map->nodes[i].pnn;
+ }
+
+ return nodes;
+}
+
uint32_t *list_of_connected_nodes(struct ctdb_context *ctdb,
struct ctdb_node_map *node_map,
TALLOC_CTX *mem_ctx,
return ret;
}
+/**
+ * check whether a transaction is active on a given db on a given node
+ */
+int32_t ctdb_ctrl_transaction_active(struct ctdb_context *ctdb,
+ uint32_t destnode,
+ uint32_t db_id)
+{
+ int32_t status;
+ int ret;
+ TDB_DATA indata;
+
+ indata.dptr = (uint8_t *)&db_id;
+ indata.dsize = sizeof(db_id);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_TRANS2_ACTIVE,
+ 0, indata, NULL, NULL, &status,
+ NULL, NULL);
+
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR, (__location__ " ctdb control for transaction_active failed\n"));
+ return -1;
+ }
+
+ return status;
+}
+
+
struct ctdb_transaction_handle {
struct ctdb_db_context *ctdb_db;
bool in_replay;
- /* we store the reads and writes done under a transaction one
- list stores both reads and writes, the other just writes
- */
+ /*
+ * we store the reads and writes done under a transaction:
+ * - one list stores both reads and writes (m_all),
+ * - the other just writes (m_write)
+ */
struct ctdb_marshall_buffer *m_all;
struct ctdb_marshall_buffer *m_write;
};
{
struct ctdb_record_handle *rh;
TDB_DATA key;
+ TDB_DATA data;
struct ctdb_ltdb_header header;
TALLOC_CTX *tmp_ctx;
const char *keyname = CTDB_TRANSACTION_LOCK_KEY;
int ret;
struct ctdb_db_context *ctdb_db = h->ctdb_db;
+ pid_t pid;
+ int32_t status;
key.dptr = discard_const(keyname);
key.dsize = strlen(keyname);
rh = ctdb_fetch_lock(ctdb_db, tmp_ctx, key, NULL);
if (rh == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to fetch_lock database\n"));
+ DEBUG(DEBUG_ERR,(__location__ " Failed to fetch_lock database\n"));
talloc_free(tmp_ctx);
return -1;
}
+
+ status = ctdb_ctrl_transaction_active(ctdb_db->ctdb,
+ CTDB_CURRENT_NODE,
+ ctdb_db->db_id);
+ if (status == 1) {
+ unsigned long int usec = (1000 + random()) % 100000;
+ DEBUG(DEBUG_DEBUG, (__location__ " transaction is active "
+ "on db_id[0x%08x]. waiting for %lu "
+ "microseconds\n",
+ ctdb_db->db_id, usec));
+ talloc_free(tmp_ctx);
+ usleep(usec);
+ goto again;
+ }
+
+ /*
+ * store the pid in the database:
+ * it is not enough that the node is dmaster...
+ */
+ pid = getpid();
+ data.dptr = (unsigned char *)&pid;
+ data.dsize = sizeof(pid_t);
+ rh->header.rsn++;
+ rh->header.dmaster = ctdb_db->ctdb->pnn;
+ ret = ctdb_ltdb_store(ctdb_db, key, &(rh->header), data);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR, (__location__ " Failed to store pid in "
+ "transaction record\n"));
+ talloc_free(tmp_ctx);
+ return -1;
+ }
+
talloc_free(rh);
ret = tdb_transaction_start(ctdb_db->ltdb->tdb);
return -1;
}
- ret = ctdb_ltdb_fetch(ctdb_db, key, &header, tmp_ctx, NULL);
- if (ret != 0 || header.dmaster != ctdb_db->ctdb->pnn) {
+ ret = ctdb_ltdb_fetch(ctdb_db, key, &header, tmp_ctx, &data);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " Failed to re-fetch transaction "
+ "lock record inside transaction\n"));
+ tdb_transaction_cancel(ctdb_db->ltdb->tdb);
+ talloc_free(tmp_ctx);
+ goto again;
+ }
+
+ if (header.dmaster != ctdb_db->ctdb->pnn) {
+ DEBUG(DEBUG_DEBUG,(__location__ " not dmaster any more on "
+ "transaction lock record\n"));
+ tdb_transaction_cancel(ctdb_db->ltdb->tdb);
+ talloc_free(tmp_ctx);
+ goto again;
+ }
+
+ if ((data.dsize != sizeof(pid_t)) || (*(pid_t *)(data.dptr) != pid)) {
+ DEBUG(DEBUG_DEBUG, (__location__ " my pid is not stored in "
+ "the transaction lock record\n"));
tdb_transaction_cancel(ctdb_db->ltdb->tdb);
talloc_free(tmp_ctx);
goto again;
&timeout, NULL);
if (ret != 0 || status != 0) {
tdb_transaction_cancel(h->ctdb_db->ltdb->tdb);
+ DEBUG(DEBUG_NOTICE, (__location__ " transaction commit%s failed"
+ ", retrying after 1 second...\n",
+ (retries==0)?"":"retry "));
sleep(1);
if (ret != 0) {
}
}
- if (++retries == 10) {
+ if (++retries == 100) {
DEBUG(DEBUG_ERR,(__location__ " Giving up transaction on db 0x%08x after %d retries failure_control=%u\n",
h->ctdb_db->db_id, retries, (unsigned)failure_control));
ctdb_control(ctdb, CTDB_CURRENT_NODE, h->ctdb_db->db_id,
}
if (ctdb_replay_transaction(h) != 0) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to replay transaction\n"));
+ DEBUG(DEBUG_ERR, (__location__ " Failed to replay "
+ "transaction on db 0x%08x, "
+ "failure control =%u\n",
+ h->ctdb_db->db_id,
+ (unsigned)failure_control));
ctdb_control(ctdb, CTDB_CURRENT_NODE, h->ctdb_db->db_id,
failure_control, CTDB_CTRL_FLAG_NOREPLY,
tdb_null, NULL, NULL, NULL, NULL, NULL);
/* do the real commit locally */
ret = tdb_transaction_commit(h->ctdb_db->ltdb->tdb);
if (ret != 0) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to commit transaction\n"));
+ DEBUG(DEBUG_ERR, (__location__ " Failed to commit transaction "
+ "on db id 0x%08x locally, "
+ "failure_control=%u\n",
+ h->ctdb_db->db_id,
+ (unsigned)failure_control));
ctdb_control(ctdb, CTDB_CURRENT_NODE, h->ctdb_db->db_id,
failure_control, CTDB_CTRL_FLAG_NOREPLY,
tdb_null, NULL, NULL, NULL, NULL, NULL);
* to the daemon as a client process, this function can be used to change
* the ctdb context from daemon into client mode
*/
-int switch_from_server_to_client(struct ctdb_context *ctdb)
+int switch_from_server_to_client(struct ctdb_context *ctdb, const char *fmt, ...)
{
int ret;
+ va_list ap;
+
+ /* Add extra information so we can identify this in the logs */
+ va_start(ap, fmt);
+ debug_extra = talloc_strdup_append(talloc_vasprintf(NULL, fmt, ap), ":");
+ va_end(ap);
/* shutdown the transport */
if (ctdb->methods) {
/* get a new event context */
talloc_free(ctdb->ev);
ctdb->ev = event_context_init(ctdb);
+ tevent_loop_allow_nesting(ctdb->ev);
close(ctdb->daemon.sd);
ctdb->daemon.sd = -1;
- /* the client does not need to be realtime */
- if (ctdb->do_setsched) {
- ctdb_restore_scheduler(ctdb);
- }
-
/* initialise ctdb */
ret = ctdb_socket_connect(ctdb);
if (ret != 0) {
}
/*
- tell the main daemon we are starting a new monitor event script
- */
-int ctdb_ctrl_event_script_init(struct ctdb_context *ctdb)
-{
- int ret;
- int32_t res;
-
- ret = ctdb_control(ctdb, CTDB_CURRENT_NODE, 0, CTDB_CONTROL_EVENT_SCRIPT_INIT, 0, tdb_null,
- ctdb, NULL, &res, NULL, NULL);
- if (ret != 0 || res != 0) {
- DEBUG(DEBUG_ERR,("Failed to send event_script_init\n"));
- return -1;
- }
-
- return 0;
-}
-
-/*
- tell the main daemon we are starting a new monitor event script
- */
-int ctdb_ctrl_event_script_finished(struct ctdb_context *ctdb)
-{
- int ret;
- int32_t res;
-
- ret = ctdb_control(ctdb, CTDB_CURRENT_NODE, 0, CTDB_CONTROL_EVENT_SCRIPT_FINISHED, 0, tdb_null,
- ctdb, NULL, &res, NULL, NULL);
- if (ret != 0 || res != 0) {
- DEBUG(DEBUG_ERR,("Failed to send event_script_init\n"));
- return -1;
- }
-
- return 0;
-}
-
-/*
- tell the main daemon we are starting to run an eventscript
- */
-int ctdb_ctrl_event_script_start(struct ctdb_context *ctdb, const char *name)
-{
- int ret;
- int32_t res;
- TDB_DATA data;
-
- data.dptr = discard_const(name);
- data.dsize = strlen(name)+1;
-
- ret = ctdb_control(ctdb, CTDB_CURRENT_NODE, 0, CTDB_CONTROL_EVENT_SCRIPT_START, 0, data,
- ctdb, NULL, &res, NULL, NULL);
- if (ret != 0 || res != 0) {
- DEBUG(DEBUG_ERR,("Failed to send event_script_start\n"));
- return -1;
- }
-
- return 0;
-}
-
-/*
- tell the main daemon the status of the script we ran
- */
-int ctdb_ctrl_event_script_stop(struct ctdb_context *ctdb, int32_t result)
-{
- int ret;
- int32_t res;
- TDB_DATA data;
-
- data.dptr = (uint8_t *)&result;
- data.dsize = sizeof(result);
-
- ret = ctdb_control(ctdb, CTDB_CURRENT_NODE, 0, CTDB_CONTROL_EVENT_SCRIPT_STOP, 0, data,
- ctdb, NULL, &res, NULL, NULL);
- if (ret != 0 || res != 0) {
- DEBUG(DEBUG_ERR,("Failed to send event_script_stop\n"));
- return -1;
- }
-
- return 0;
-}
-
-
-/*
- get the status of running the monitor eventscripts
+ get the status of running the monitor eventscripts: NULL means never run.
*/
int ctdb_ctrl_getscriptstatus(struct ctdb_context *ctdb,
struct timeval timeout, uint32_t destnode,
- TALLOC_CTX *mem_ctx,
- struct ctdb_monitoring_wire **script_status)
+ TALLOC_CTX *mem_ctx, enum ctdb_eventscript_call type,
+ struct ctdb_scripts_wire **script_status)
{
int ret;
- TDB_DATA outdata;
+ TDB_DATA outdata, indata;
int32_t res;
+ uint32_t uinttype = type;
+
+ indata.dptr = (uint8_t *)&uinttype;
+ indata.dsize = sizeof(uinttype);
ret = ctdb_control(ctdb, destnode, 0,
- CTDB_CONTROL_GET_EVENT_SCRIPT_STATUS, 0, tdb_null,
+ CTDB_CONTROL_GET_EVENT_SCRIPT_STATUS, 0, indata,
mem_ctx, &outdata, &res, &timeout, NULL);
- if (ret != 0 || res != 0 || outdata.dsize == 0) {
+ if (ret != 0 || res != 0) {
DEBUG(DEBUG_ERR,(__location__ " ctdb_control for getscriptstatus failed ret:%d res:%d\n", ret, res));
return -1;
}
- *script_status = (struct ctdb_monitoring_wire *)talloc_memdup(mem_ctx, outdata.dptr, outdata.dsize);
- talloc_free(outdata.dptr);
+ if (outdata.dsize == 0) {
+ *script_status = NULL;
+ } else {
+ *script_status = (struct ctdb_scripts_wire *)talloc_memdup(mem_ctx, outdata.dptr, outdata.dsize);
+ talloc_free(outdata.dptr);
+ }
return 0;
}
return 0;
}
+
+/* enable an eventscript
+ */
+int ctdb_ctrl_enablescript(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, const char *script)
+{
+ int ret;
+ TDB_DATA data;
+ int32_t res;
+
+ data.dsize = strlen(script) + 1;
+ data.dptr = discard_const(script);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_ENABLE_SCRIPT, 0, data,
+ NULL, NULL, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for enablescript failed\n"));
+ return -1;
+ }
+
+ return 0;
+}
+
+/* disable an eventscript
+ */
+int ctdb_ctrl_disablescript(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, const char *script)
+{
+ int ret;
+ TDB_DATA data;
+ int32_t res;
+
+ data.dsize = strlen(script) + 1;
+ data.dptr = discard_const(script);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_DISABLE_SCRIPT, 0, data,
+ NULL, NULL, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for disablescript failed\n"));
+ return -1;
+ }
+
+ return 0;
+}
+
+
+int ctdb_ctrl_set_ban(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, struct ctdb_ban_time *bantime)
+{
+ int ret;
+ TDB_DATA data;
+ int32_t res;
+
+ data.dsize = sizeof(*bantime);
+ data.dptr = (uint8_t *)bantime;
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_SET_BAN_STATE, 0, data,
+ NULL, NULL, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for set ban state failed\n"));
+ return -1;
+ }
+
+ return 0;
+}
+
+
+int ctdb_ctrl_get_ban(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, TALLOC_CTX *mem_ctx, struct ctdb_ban_time **bantime)
+{
+ int ret;
+ TDB_DATA outdata;
+ int32_t res;
+ TALLOC_CTX *tmp_ctx = talloc_new(NULL);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_GET_BAN_STATE, 0, tdb_null,
+ tmp_ctx, &outdata, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for set ban state failed\n"));
+ talloc_free(tmp_ctx);
+ return -1;
+ }
+
+ *bantime = (struct ctdb_ban_time *)talloc_steal(mem_ctx, outdata.dptr);
+ talloc_free(tmp_ctx);
+
+ return 0;
+}
+
+
+int ctdb_ctrl_set_db_priority(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, struct ctdb_db_priority *db_prio)
+{
+ int ret;
+ int32_t res;
+ TDB_DATA data;
+ TALLOC_CTX *tmp_ctx = talloc_new(NULL);
+
+ data.dptr = (uint8_t*)db_prio;
+ data.dsize = sizeof(*db_prio);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_SET_DB_PRIORITY, 0, data,
+ tmp_ctx, NULL, &res, &timeout, NULL);
+ if (ret != 0 || res != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for set_db_priority failed\n"));
+ talloc_free(tmp_ctx);
+ return -1;
+ }
+
+ talloc_free(tmp_ctx);
+
+ return 0;
+}
+
+int ctdb_ctrl_get_db_priority(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, uint32_t db_id, uint32_t *priority)
+{
+ int ret;
+ int32_t res;
+ TDB_DATA data;
+ TALLOC_CTX *tmp_ctx = talloc_new(NULL);
+
+ data.dptr = (uint8_t*)&db_id;
+ data.dsize = sizeof(db_id);
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_GET_DB_PRIORITY, 0, data,
+ tmp_ctx, NULL, &res, &timeout, NULL);
+ if (ret != 0 || res < 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for set_db_priority failed\n"));
+ talloc_free(tmp_ctx);
+ return -1;
+ }
+
+ if (priority) {
+ *priority = res;
+ }
+
+ talloc_free(tmp_ctx);
+
+ return 0;
+}
+
+int ctdb_ctrl_getstathistory(struct ctdb_context *ctdb, struct timeval timeout, uint32_t destnode, TALLOC_CTX *mem_ctx, struct ctdb_statistics_wire **stats)
+{
+ int ret;
+ TDB_DATA outdata;
+ int32_t res;
+
+ ret = ctdb_control(ctdb, destnode, 0,
+ CTDB_CONTROL_GET_STAT_HISTORY, 0, tdb_null,
+ mem_ctx, &outdata, &res, &timeout, NULL);
+ if (ret != 0 || res != 0 || outdata.dsize == 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_control for getstathistory failed ret:%d res:%d\n", ret, res));
+ return -1;
+ }
+
+ *stats = (struct ctdb_statistics_wire *)talloc_memdup(mem_ctx, outdata.dptr, outdata.dsize);
+ talloc_free(outdata.dptr);
+
+ return 0;
+}
git.samba.org / sahlberg / ctdb.git / blobdiff