return 0;
}
-#if 0
-/**
- * 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 (m_all),
- * - the other just writes (m_write)
- */
- struct ctdb_marshall_buffer *m_all;
- struct ctdb_marshall_buffer *m_write;
-};
-
-/* start a transaction on a database */
-static int ctdb_transaction_destructor(struct ctdb_transaction_handle *h)
-{
- tdb_transaction_cancel(h->ctdb_db->ltdb->tdb);
- return 0;
-}
-
-/* start a transaction on a database */
-static int ctdb_transaction_fetch_start(struct ctdb_transaction_handle *h)
-{
- 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);
-
- if (!ctdb_db->persistent) {
- DEBUG(DEBUG_ERR,(__location__ " Attempted transaction on non-persistent database\n"));
- return -1;
- }
-
-again:
- tmp_ctx = talloc_new(h);
-
- rh = ctdb_fetch_lock(ctdb_db, tmp_ctx, key, NULL);
- if (rh == NULL) {
- 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);
- if (ret != 0) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to start tdb transaction\n"));
- talloc_free(tmp_ctx);
- return -1;
- }
-
- 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;
- }
-
- talloc_free(tmp_ctx);
-
- return 0;
-}
-
-
-/* start a transaction on a database */
-struct ctdb_transaction_handle *ctdb_transaction_start(struct ctdb_db_context *ctdb_db,
- TALLOC_CTX *mem_ctx)
-{
- struct ctdb_transaction_handle *h;
- int ret;
-
- h = talloc_zero(mem_ctx, struct ctdb_transaction_handle);
- if (h == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " oom for transaction handle\n"));
- return NULL;
- }
-
- h->ctdb_db = ctdb_db;
-
- ret = ctdb_transaction_fetch_start(h);
- if (ret != 0) {
- talloc_free(h);
- return NULL;
- }
-
- talloc_set_destructor(h, ctdb_transaction_destructor);
-
- return h;
-}
-
-
-
-/*
- fetch a record inside a transaction
- */
-int ctdb_transaction_fetch(struct ctdb_transaction_handle *h,
- TALLOC_CTX *mem_ctx,
- TDB_DATA key, TDB_DATA *data)
-{
- struct ctdb_ltdb_header header;
- int ret;
-
- ZERO_STRUCT(header);
-
- ret = ctdb_ltdb_fetch(h->ctdb_db, key, &header, mem_ctx, data);
- if (ret == -1 && header.dmaster == (uint32_t)-1) {
- /* record doesn't exist yet */
- *data = tdb_null;
- ret = 0;
- }
-
- if (ret != 0) {
- return ret;
- }
-
- if (!h->in_replay) {
- h->m_all = ctdb_marshall_add(h, h->m_all, h->ctdb_db->db_id, 1, key, NULL, *data);
- if (h->m_all == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to add to marshalling record\n"));
- return -1;
- }
- }
-
- return 0;
-}
-
-/*
- stores a record inside a transaction
- */
-int ctdb_transaction_store(struct ctdb_transaction_handle *h,
- TDB_DATA key, TDB_DATA data)
-{
- TALLOC_CTX *tmp_ctx = talloc_new(h);
- struct ctdb_ltdb_header header;
- TDB_DATA olddata;
- int ret;
-
- ZERO_STRUCT(header);
-
- /* we need the header so we can update the RSN */
- ret = ctdb_ltdb_fetch(h->ctdb_db, key, &header, tmp_ctx, &olddata);
- if (ret == -1 && header.dmaster == (uint32_t)-1) {
- /* the record doesn't exist - create one with us as dmaster.
- This is only safe because we are in a transaction and this
- is a persistent database */
- ZERO_STRUCT(header);
- } else if (ret != 0) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to fetch record\n"));
- talloc_free(tmp_ctx);
- return ret;
- }
-
- if (data.dsize == olddata.dsize &&
- memcmp(data.dptr, olddata.dptr, data.dsize) == 0) {
- /* save writing the same data */
- talloc_free(tmp_ctx);
- return 0;
- }
-
- header.dmaster = h->ctdb_db->ctdb->pnn;
- header.rsn++;
-
- if (!h->in_replay) {
- h->m_all = ctdb_marshall_add(h, h->m_all, h->ctdb_db->db_id, 0, key, NULL, data);
- if (h->m_all == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to add to marshalling record\n"));
- talloc_free(tmp_ctx);
- return -1;
- }
- }
-
- h->m_write = ctdb_marshall_add(h, h->m_write, h->ctdb_db->db_id, 0, key, &header, data);
- if (h->m_write == NULL) {
- DEBUG(DEBUG_ERR,(__location__ " Failed to add to marshalling record\n"));
- talloc_free(tmp_ctx);
- return -1;
- }
-
- ret = ctdb_ltdb_store(h->ctdb_db, key, &header, data);
-
- talloc_free(tmp_ctx);
-
- return ret;
-}
-
-/*
- replay a transaction
- */
-static int ctdb_replay_transaction(struct ctdb_transaction_handle *h)
-{
- int ret, i;
- struct ctdb_rec_data *rec = NULL;
-
- h->in_replay = true;
- talloc_free(h->m_write);
- h->m_write = NULL;
-
- ret = ctdb_transaction_fetch_start(h);
- if (ret != 0) {
- return ret;
- }
-
- for (i=0;i<h->m_all->count;i++) {
- TDB_DATA key, data;
-
- rec = ctdb_marshall_loop_next(h->m_all, rec, NULL, NULL, &key, &data);
- if (rec == NULL) {
- DEBUG(DEBUG_ERR, (__location__ " Out of records in ctdb_replay_transaction?\n"));
- goto failed;
- }
-
- if (rec->reqid == 0) {
- /* its a store */
- if (ctdb_transaction_store(h, key, data) != 0) {
- goto failed;
- }
- } else {
- TDB_DATA data2;
- TALLOC_CTX *tmp_ctx = talloc_new(h);
-
- if (ctdb_transaction_fetch(h, tmp_ctx, key, &data2) != 0) {
- talloc_free(tmp_ctx);
- goto failed;
- }
- if (data2.dsize != data.dsize ||
- memcmp(data2.dptr, data.dptr, data.dsize) != 0) {
- /* the record has changed on us - we have to give up */
- talloc_free(tmp_ctx);
- goto failed;
- }
- talloc_free(tmp_ctx);
- }
- }
-
- return 0;
-
-failed:
- tdb_transaction_cancel(h->ctdb_db->ltdb->tdb);
- return -1;
-}
-
-
-/*
- commit a transaction
- */
-int ctdb_transaction_commit(struct ctdb_transaction_handle *h)
-{
- int ret, retries=0;
- int32_t status;
- struct ctdb_context *ctdb = h->ctdb_db->ctdb;
- struct timeval timeout;
- enum ctdb_controls failure_control = CTDB_CONTROL_TRANS2_ERROR;
-
- talloc_set_destructor(h, NULL);
-
- /* our commit strategy is quite complex.
-
- - we first try to commit the changes to all other nodes
-
- - if that works, then we commit locally and we are done
-
- - if a commit on another node fails, then we need to cancel
- the transaction, then restart the transaction (thus
- opening a window of time for a pending recovery to
- complete), then replay the transaction, checking all the
- reads and writes (checking that reads give the same data,
- and writes succeed). Then we retry the transaction to the
- other nodes
- */
-
-again:
- if (h->m_write == NULL) {
- /* no changes were made */
- tdb_transaction_cancel(h->ctdb_db->ltdb->tdb);
- talloc_free(h);
- return 0;
- }
-
- /* tell ctdbd to commit to the other nodes */
- timeout = timeval_current_ofs(1, 0);
- ret = ctdb_control(ctdb, CTDB_CURRENT_NODE, h->ctdb_db->db_id,
- retries==0?CTDB_CONTROL_TRANS2_COMMIT:CTDB_CONTROL_TRANS2_COMMIT_RETRY, 0,
- ctdb_marshall_finish(h->m_write), NULL, NULL, &status,
- &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) {
- failure_control = CTDB_CONTROL_TRANS2_ERROR;
- } else {
- /* work out what error code we will give if we
- have to fail the operation */
- switch ((enum ctdb_trans2_commit_error)status) {
- case CTDB_TRANS2_COMMIT_SUCCESS:
- case CTDB_TRANS2_COMMIT_SOMEFAIL:
- case CTDB_TRANS2_COMMIT_TIMEOUT:
- failure_control = CTDB_CONTROL_TRANS2_ERROR;
- break;
- case CTDB_TRANS2_COMMIT_ALLFAIL:
- failure_control = CTDB_CONTROL_TRANS2_FINISHED;
- break;
- }
- }
-
- 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,
- failure_control, CTDB_CTRL_FLAG_NOREPLY,
- tdb_null, NULL, NULL, NULL, NULL, NULL);
- talloc_free(h);
- return -1;
- }
-
- if (ctdb_replay_transaction(h) != 0) {
- 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);
- talloc_free(h);
- return -1;
- }
- goto again;
- } else {
- failure_control = CTDB_CONTROL_TRANS2_ERROR;
- }
-
- /* 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 "
- "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);
- talloc_free(h);
- return ret;
- }
-
- /* tell ctdbd that we are finished with our local commit */
- ctdb_control(ctdb, CTDB_CURRENT_NODE, h->ctdb_db->db_id,
- CTDB_CONTROL_TRANS2_FINISHED, CTDB_CTRL_FLAG_NOREPLY,
- tdb_null, NULL, NULL, NULL, NULL, NULL);
- talloc_free(h);
- return 0;
-}
-#endif
/*
recovery daemon ping to main daemon
git.samba.org / ctdb.git / commitdiff