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

Update MariaDBMonitor documentation

Browse Source 
Mention the secondary use of switchover_timeout. Formatting fixes.
This commit is contained in:
Esa Korhonen 2018-11-07 13:20:02 +02:00
parent 07d497a672
commit 56d28d703c
1 changed files with 33 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

67
Documentation/Monitors/MariaDB-Monitor.md
View File

@ -299,37 +299,34 @@ following:
1. Select the most up-to-date slave of the old master to be the new master. The
selection criteria is as follows in descending priority:
1. gtid_IO_pos (latest event in relay log)
2. gtid_current_pos (most processed events)
3. log_slave_updates is on
4. disk space is not low
1. gtid_IO_pos (latest event in relay log)
2. gtid_current_pos (most processed events)
3. log_slave_updates is on
4. disk space is not low
2. If the new master has unprocessed relay log items, cancel and try again
later.
3. Prepare the new master:
1. Remove the slave connection the new master used to replicate from the old
1. Remove the slave connection the new master used to replicate from the old
master.
2. Disable the *read\_only*-flag.
3. Enable scheduled server events (if event handling is on).
4. Run the commands in `promotion_sql_file`.
5. Start replication from external master is one existed.
2. Disable the *read\_only*-flag.
3. Enable scheduled server events (if event handling is on).
4. Run the commands in `promotion_sql_file`.
5. Start replication from external master if one existed.
4. Redirect all other slaves to replicate from the new master:
1. STOP SLAVE and RESET SLAVE
2. CHANGE MASTER TO
3. START SLAVE
1. STOP SLAVE and RESET SLAVE
2. CHANGE MASTER TO
3. START SLAVE
5. Check that all slaves are replicating.
Failover may lose events if no slave managed to replicate the events before the
master went down.
**Switchover** swaps a running master with a running slave. It does the
following:
1. Prepare the old master for demotion:
1. Stop any external replication.
2. Enable the *read\_only*-flag to stop writes.
3. Disable scheduled server events (if event handling is on).
4. Run the commands in `demotion_sql_file`.
5. Flush the binary log (FLUSH LOGS) so that all events are on disk.
1. Stop any external replication.
2. Enable the *read\_only*-flag to stop writes.
3. Disable scheduled server events (if event handling is on).
4. Run the commands in `demotion_sql_file`.
5. Flush the binary log (FLUSH LOGS) so that all events are on disk.
2. Wait for the new master to catch up with the old master.
3. Promote new master and redirect slaves as in failover steps 3 and 4. Also
redirect the demoted old master.
@ -353,15 +350,15 @@ cluster are out of sync while the actual data is known to be in sync. The
operation proceeds as follows:
1. Reset gtid:s and delete binary logs on all servers:
1. Stop (STOP SLAVE) and delete (RESET SLAVE ALL) all slave connections.
2. Enable the *read\_only*-flag.
3. Disable scheduled server events (if event handling is on).
3. Delete binary logs (RESET MASTER).
4. Set the sequence number of *gtid\_slave\_pos* to zero. This also affects
1. Stop (STOP SLAVE) and delete (RESET SLAVE ALL) all slave connections.
2. Enable the *read\_only*-flag.
3. Disable scheduled server events (if event handling is on).
3. Delete binary logs (RESET MASTER).
4. Set the sequence number of *gtid\_slave\_pos* to zero. This also affects
*gtid\_current\_pos*.
2. Prepare new master:
1. Disable the *read\_only*-flag.
2. Enable scheduled server events (if event handling is on).
1. Disable the *read\_only*-flag.
2. Enable scheduled server events (if event handling is on).
3. Direct other servers to replicate from the new master as in the other
operations.
@ -492,6 +489,11 @@ The backends must all use GTID-based replication, and the domain id should not
change during a switchover or failover. Master and slaves must have
well-behaving GTIDs with no extra events on slave servers.
Failover cannot be performed if MaxScale was started only after the master
server went down. This is because MaxScale needs reliable information on the
gtid domain of the cluster and the replication topology in general to properly
select the new master.
Failover may lose events. If a master goes down before sending new events to at
least one slave, those events are lost when a new master is chosen. If the old
master comes back online, the other servers have likely moved on with a
@ -614,18 +616,15 @@ encrypted with the same key to avoid erroneous decryption.
#### `failover_timeout` and `switchover_timeout`
Time limit for the cluster failover and switchover in seconds. The default values
are 90 seconds.
Time limit for failover and switchover operations, in seconds. The default
values are 90 seconds for both. `switchover_timeout` is also used as the time
limit for a rejoin operation. Rejoin should rarely time out, since it is a
faster operation than switchover.
If no successful failover/switchover takes place within the configured time
period, a message is logged and automatic failover is disabled. This prevents
further automatic modifications to the misbehaving cluster.
`failover_timeout` also controls how long a MaxScale instance that has
transitioned from passive to active will wait for a failover to take place after
an apparent loss of a master server. If no new master server is detected within
the configured time period, failover will be initiated again.
#### `verify_master_failure` and `master_failure_timeout`
Enable additional master failure verification for automatic failover.