hcq/MaxScale
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Relay log clear supports multiple slave connections

Browse Source 
Now waits for the relay log of the correct slave connection.
This commit is contained in:
Esa Korhonen
2018-08-28 13:54:53 +03:00
parent 85d8a85cde
commit c39177bc8d
5 changed files with 57 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

51
server/modules/monitor/mariadbmon/mariadbserver.cc
View File

@ -73,15 +73,19 @@ void NodeData::reset_indexes()
in_stack = false;
}
int64_t MariaDBServer::relay_log_events()
int64_t MariaDBServer::relay_log_events(const MariaDBServer* master)
{
/* The events_ahead-call below ignores domains where current_pos is ahead of io_pos. This situation is
* rare but is possible (I guess?) if the server is replicating a domain from multiple masters
* and decides to process events from one relay log before getting new events to the other. In
* any case, such events are obsolete and the server can be considered to have processed such logs. */
// TODO: Fix for multisource repl
return !m_slave_status.empty() ? GtidList::events_ahead(m_slave_status[0].gtid_io_pos, m_gtid_current_pos,
GtidList::MISSING_DOMAIN_LHS_ADD) : 0;
int64_t rval = -1;
const SlaveStatus* sstatus = slave_connection_status(master);
if (sstatus)
{
rval = sstatus->gtid_io_pos.events_ahead(m_gtid_current_pos, GtidList::MISSING_DOMAIN_IGNORE);
}
return rval;
}
std::unique_ptr<QueryResult> MariaDBServer::execute_query(const string& query, std::string* errmsg_out)
@ -397,7 +401,7 @@ bool MariaDBServer::wait_until_gtid(const GtidList& target, int timeout, json_t*
if (update_gtids())
{
const GtidList& compare_to = use_binlog_pos ? m_gtid_binlog_pos : m_gtid_current_pos;
if (GtidList::events_ahead(target, compare_to, GtidList::MISSING_DOMAIN_IGNORE) == 0)
if (target.events_ahead(compare_to, GtidList::MISSING_DOMAIN_IGNORE) == 0)
{
gtid_reached = true;
}
@ -656,31 +660,37 @@ bool MariaDBServer::join_cluster(const string& change_cmd)
return success;
}
bool MariaDBServer::failover_wait_relay_log(int seconds_remaining, json_t** err_out)
bool MariaDBServer::failover_wait_relay_log(const MariaDBServer* master, int seconds_remaining,
json_t** err_out)
{
time_t begin = time(NULL);
bool query_ok = true;
bool io_pos_stable = true;
while (relay_log_events() > 0 &&
query_ok &&
io_pos_stable &&
difftime(time(NULL), begin) < seconds_remaining)
int64_t events = relay_log_events(master);
while (events > 0 && query_ok && io_pos_stable && difftime(time(NULL), begin) < seconds_remaining)
{
MXS_INFO("Relay log of server '%s' not yet empty, waiting to clear %" PRId64 " events.",
name(), relay_log_events());
const SlaveStatus* sstatus = slave_connection_status(master);
mxb_assert(sstatus);
GtidList old_gtid_io_pos = sstatus->gtid_io_pos;
// Sleep for a while before querying server again.
std::this_thread::sleep_for(std::chrono::milliseconds(1000));
// TODO: check server version before entering failover.
// TODO: fix for multisource
GtidList old_gtid_io_pos = m_slave_status[0].gtid_io_pos;
MXS_NOTICE("Relay log of server '%s' not yet empty, waiting to clear %" PRId64 " events.",
name(), events);
std::this_thread::sleep_for(std::chrono::seconds(1));
// Update gtid:s first to make sure Gtid_IO_Pos is the more recent value.
// It doesn't matter here, but is a general rule.
query_ok = update_gtids() && do_show_slave_status();
io_pos_stable = (old_gtid_io_pos == m_slave_status[0].gtid_io_pos);
if (query_ok)
{
const SlaveStatus* new_sstatus = slave_connection_status(master);
io_pos_stable = new_sstatus ? (old_gtid_io_pos == new_sstatus->gtid_io_pos) : false;
events = relay_log_events(master);
}
}
bool rval = false;
if (relay_log_events() == 0)
if (events == 0 && query_ok && io_pos_stable)
{
rval = true;
}
@ -695,10 +705,9 @@ bool MariaDBServer::failover_wait_relay_log(int seconds_remaining, json_t** err_
{
reason = "Old master sent new event(s)";
}
else if (relay_log_events() < 0) // TODO: This is currently impossible
else if (events < 0)
{
reason = "Invalid Gtid(s) (current_pos: " + m_gtid_current_pos.to_string() +
", io_pos: " + m_slave_status[0].gtid_io_pos.to_string() + ")";
reason = string_printf("Slave connection to '%s' was removed", master->name());
}
PRINT_MXS_JSON_ERROR(err_out, "Failover: %s while waiting for server '%s' to process relay log. "
"Cancelling failover.", reason.c_str(), name());