protocol design and packet details
*/
#include "includes.h"
-#include "lib/events/events.h"
+#include "lib/tevent/tevent.h"
#include "lib/tdb/include/tdb.h"
#include "lib/util/dlinklist.h"
#include "system/network.h"
int msglen, len;
if (ctdb->methods == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to send error. Transport is DOWN\n"));
+ DEBUG(DEBUG_INFO,(__location__ " Failed to send error. Transport is DOWN\n"));
return;
}
}
if (ctdb->methods == NULL) {
- ctdb_fatal(ctdb, "ctdb_send_dmaster_reply cant update dmaster sicne transport is down");
+ ctdb_fatal(ctdb, "ctdb_send_dmaster_reply cant update dmaster since transport is down");
return;
}
struct ctdb_call_state *state;
struct ctdb_context *ctdb = ctdb_db->ctdb;
struct ctdb_ltdb_header header;
+ int ret;
- DEBUG(DEBUG_INFO,("pnn %u dmaster response %08x\n", ctdb->pnn, ctdb_hash(&key)));
+ DEBUG(DEBUG_DEBUG,("pnn %u dmaster response %08x\n", ctdb->pnn, ctdb_hash(&key)));
ZERO_STRUCT(header);
header.rsn = rsn + 1;
if (ctdb_ltdb_store(ctdb_db, key, &header, data) != 0) {
ctdb_fatal(ctdb, "ctdb_reply_dmaster store failed\n");
- ctdb_ltdb_unlock(ctdb_db, key);
+
+ ret = ctdb_ltdb_unlock(ctdb_db, key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
return;
}
if (state == NULL) {
DEBUG(DEBUG_ERR,("pnn %u Invalid reqid %u in ctdb_become_dmaster from node %u\n",
ctdb->pnn, hdr->reqid, hdr->srcnode));
- ctdb_ltdb_unlock(ctdb_db, key);
+
+ ret = ctdb_ltdb_unlock(ctdb_db, key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
+ return;
+ }
+
+ if (key.dsize != state->call->key.dsize || memcmp(key.dptr, state->call->key.dptr, key.dsize)) {
+ DEBUG(DEBUG_ERR, ("Got bogus DMASTER packet reqid:%u from node %u. Key does not match key held in matching idr.\n", hdr->reqid, hdr->srcnode));
+
+ ret = ctdb_ltdb_unlock(ctdb_db, key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
return;
}
if (hdr->reqid != state->reqid) {
/* we found a record but it was the wrong one */
DEBUG(DEBUG_ERR, ("Dropped orphan in ctdb_become_dmaster with reqid:%u\n from node %u", hdr->reqid, hdr->srcnode));
- ctdb_ltdb_unlock(ctdb_db, key);
+
+ ret = ctdb_ltdb_unlock(ctdb_db, key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
return;
}
- ctdb_call_local(ctdb_db, state->call, &header, state, &data, ctdb->pnn);
+ ctdb_call_local(ctdb_db, state->call, &header, state, &data);
- ctdb_ltdb_unlock(ctdb_db, state->call->key);
+ ret = ctdb_ltdb_unlock(ctdb_db, state->call->key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
state->state = CTDB_CALL_DONE;
if (state->async.fn) {
ctdb_fatal(ctdb, "ctdb_req_dmaster to non-lmaster");
}
- DEBUG(DEBUG_INFO,("pnn %u dmaster request on %08x for %u from %u\n",
+ DEBUG(DEBUG_DEBUG,("pnn %u dmaster request on %08x for %u from %u\n",
ctdb->pnn, ctdb_hash(&key), c->dmaster, c->hdr.srcnode));
/* its a protocol error if the sending node is not the current dmaster */
ctdb_db->db_id, hdr->generation, ctdb->vnn_map->generation,
(unsigned long long)c->rsn, (unsigned long long)header.rsn, c->hdr.reqid,
(key.dsize >= 4)?(*(uint32_t *)key.dptr):0));
- /*
- * with the new vacuuming code there are conditions where a node has outdated
- * information about the real dmaster
- * since here we are lmaster and always know who is the real dmaster
- * we don't need to exit with a fatal error and we even don't have
- * to initiate a recovery
- */
+ if (header.rsn != 0 || header.dmaster != ctdb->pnn) {
+ DEBUG(DEBUG_ERR,("ctdb_req_dmaster from non-master. Force a recovery.\n"));
+
+ ctdb->recovery_mode = CTDB_RECOVERY_ACTIVE;
+ return;
+ }
}
if (header.rsn > c->rsn) {
ctdb_become_dmaster(ctdb_db, hdr, key, data, c->rsn);
} else {
ctdb_send_dmaster_reply(ctdb_db, &header, key, data, c->dmaster, hdr->reqid);
- ctdb_ltdb_unlock(ctdb_db, key);
+
+ ret = ctdb_ltdb_unlock(ctdb_db, key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
}
}
struct ctdb_db_context *ctdb_db;
if (ctdb->methods == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed ctdb_request_call. Transport is DOWN\n"));
+ DEBUG(DEBUG_INFO,(__location__ " Failed ctdb_request_call. Transport is DOWN\n"));
return;
}
if (header.dmaster != ctdb->pnn) {
talloc_free(data.dptr);
ctdb_call_send_redirect(ctdb, call->key, c, &header);
- ctdb_ltdb_unlock(ctdb_db, call->key);
+
+ ret = ctdb_ltdb_unlock(ctdb_db, call->key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
return;
}
- if (c->hopcount > ctdb->statistics.max_hop_count) {
- ctdb->statistics.max_hop_count = c->hopcount;
- }
+ CTDB_UPDATE_STAT(ctdb, max_hop_count, c->hopcount);
- /* if this nodes has done enough consecutive calls on the same record
- then give them the record
- or if the node requested an immediate migration
- */
- if ( c->hdr.srcnode != ctdb->pnn &&
- ((header.laccessor == c->hdr.srcnode
- && header.lacount >= ctdb->tunable.max_lacount)
- || (c->flags & CTDB_IMMEDIATE_MIGRATION)) ) {
+ /* Try if possible to migrate the record off to the caller node.
+ * From the clients perspective a fetch of the data is just as
+ * expensive as a migration.
+ */
+ if (c->hdr.srcnode != ctdb->pnn) {
if (ctdb_db->transaction_active) {
- DEBUG(DEBUG_WARNING, (__location__ " refusing migration"
+ DEBUG(DEBUG_INFO, (__location__ " refusing migration"
" of key %s while transaction is active\n",
(char *)call->key.dptr));
} else {
- DEBUG(DEBUG_INFO,("pnn %u starting migration of %08x to %u\n",
+ DEBUG(DEBUG_DEBUG,("pnn %u starting migration of %08x to %u\n",
ctdb->pnn, ctdb_hash(&(call->key)), c->hdr.srcnode));
ctdb_call_send_dmaster(ctdb_db, c, &header, &(call->key), &data);
talloc_free(data.dptr);
- ctdb_ltdb_unlock(ctdb_db, call->key);
+
+ ret = ctdb_ltdb_unlock(ctdb_db, call->key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
return;
}
}
- ctdb_call_local(ctdb_db, call, &header, hdr, &data, c->hdr.srcnode);
+ ctdb_call_local(ctdb_db, call, &header, hdr, &data);
- ctdb_ltdb_unlock(ctdb_db, call->key);
+ ret = ctdb_ltdb_unlock(ctdb_db, call->key);
+ if (ret != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " ctdb_ltdb_unlock() failed with error %d\n", ret));
+ }
len = offsetof(struct ctdb_reply_call, data) + call->reply_data.dsize;
r = ctdb_transport_allocate(ctdb, ctdb, CTDB_REPLY_CALL, len,
*(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);
event_add_timed(ctdb->ev, state, timeval_zero(), call_local_trigger, state);
struct ctdb_context *ctdb = ctdb_db->ctdb;
if (ctdb->methods == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed send packet. Transport is down\n"));
+ DEBUG(DEBUG_INFO,(__location__ " Failed send packet. Transport is down\n"));
return NULL;
}
struct ctdb_req_keepalive *r;
if (ctdb->methods == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to send keepalive. Transport is DOWN\n"));
+ DEBUG(DEBUG_INFO,(__location__ " Failed to send keepalive. Transport is DOWN\n"));
return;
}
r->hdr.destnode = destnode;
r->hdr.reqid = 0;
- ctdb->statistics.keepalive_packets_sent++;
+ CTDB_INCREMENT_STAT(ctdb, keepalive_packets_sent);
ctdb_queue_packet(ctdb, &r->hdr);
git.samba.org / sahlberg / ctdb.git / blobdiff