So far Patroni was populating pg_hba.conf only when running bootstrap code and after that it was not very handy to manage it's content, because it was necessary to login to every node, change pg_hba.conf manually and run pg_ctl reload.
This commit intends to fix it and give Patroni control over pg_hba.conf. It is possible to define pg_hba.conf content via `postgresql.pg_hba` in the patroni configuration file or in the `DCS/config` (dynamic configuration).
If the `hba_file` is defined in the `postgresql.parameters`, Patroni will ignore `postgresql.pg_hba`.
For backward compatibility this feature is not enabled by default. To enable it you have to set `postgresql.use_unix_socket: true`.
If feature is enable, and `unix_socket_directories` is defined and non empty, Patroni will use the first suitable value from it to connect to the local postgres cluster.
If the `unix_socket_directories` is not defined, Patroni will assume that default value should be used and will not pass `host` to command line arguments and omit it from connection url.
Solves: https://github.com/zalando/patroni/issues/61
In addition to mentioned above, this commit solves couple of bugs:
* manual failover with pg_rewind in a pause state was broken
* psycopg2 (or libpq, I am not really sure what exactly) doesn't mark cursors connection as closed when we use unix socket and there is an `OperationalError` occurs. We will close such connection on our own.
because `boto.exception` is not an excpetion, but a python module.
+ increase retry timeout to 5 minutes
+ refactor unit-tests to cover the case with retries.
pg_controldata output depends on postgres major version and in some cases some of the parameters are prefixed by 'Current ' for old postgres versions.
Bug was introduced by commit 37c1552.
Fixes https://github.com/zalando/patroni/issues/455
Previously we were running pg_rewind only in limited amount of cases:
* when we knew postgres was a master (no recovery.conf in data dir)
* when we were doing a manual switchover to a specific node (no
guaranty that this node is the most up-to-date)
* when a given node has nofailover tag (it could be ahead of new master)
This approach was kind of working in most of the cases, but sometimes we
were executing pg_rewind when it was not necessary and in some other
cases we were not executing it although it was needed.
The main idea of this PR is first try to figure out that we really need
to run pg_rewind by analyzing timelineid, LSN and history file on master
and replica and run it only if it's needed.
Current UI to change cluster configuration is somewhat unfriendly, involving a curl command, knowing the REST API endpoint, knowing the specific syntax to call it with and writing a JSON document. I added two commands in this branch to make this a bit easier, `show-config` and `edit-config` (names are merely placeholders, any opinions on better ones?).
* `patronictl show-config clustername` fetches the config from DCS, formats it as YAML and outputs it.
* `patronictl edit-config clustername` fetches the config, formats it as YAML, invokes $EDITOR on it, then shows user the diff and after confirmation applies the changed config to DCS, guarding for concurrent modifications.
* `patronictl edit-config clustername --set synchronous_mode=true --set postgresql.use_slots=true` will set the specific key-value pairs.
There are also some UI capabilities I'm less sure of, but included them here as I already implemented them.
* If output is a tty then the diffs are colored. I'm not sure if this feature is cool enough to pull the weight of adding a dependency on cdiff. Or maybe someone knows of another more task focused diff coloring library?
* `patronictl edit-config clustername --pg work_mem=100MB` - Shorthand for `--set postgresql.parameters.work_mem=100MB`
* `patronictl edit-config clustername --apply changes.yaml` - apply changes from a yaml file.
* `patronictl edit-config clustername --replace new-config.yaml` - replace config with new version.
When all etcd servers refuse connections during watch the call will fail with an exception and will be immediately retried. This creates a huge amount of log spam potentially creating additional issues on top of losing the DCS. This patch takes note if etcd failures are repeating and starting from the second failure will sleep for a second before retrying. It additionally omits the stack trace after the first failure in a streak of failures.
Default value of wal_sender_timeout is 60 seconds while we are trying to remove replication slot after 30 seconds (ttl=30). That means postgres might think that slot is still active and does nothing. Patroni at the same time was thinking that it was removed successfully.
If the drop replication slot query didn't return any single row we must fetch list of existing physical replication slots from postgres on the next iteration of HA loop.
Fixes: issue #425
* Reassemble postgresql parameters when major version became known
Otherwise we were writing some "unknown" parameters into postgresql.conf
and postgres was refusing to start. Only 9.3 was affected.
In addition to that move rename of wal_level from hot_standby to replica
into get_server_parameters method. Now this rename is handled in a
single place.
* Bump etcd and consul versions
Change hostnames by ip addresses was causing certificate verification to
fail. Instead of doing it we will better monkey patch urllib3
functionality which does name resolution. It should work without
problems even for https connection.
Previously pg_ctl waited for a timeout and then happily trodded on considering PostgreSQL to be running. This caused PostgreSQL to show up in listings as running when it was actually not and caused a race condition that resulted in either a failover or a crash recovery or a crash recovery interrupted by failover and a missed rewind.
This change adds a master_start_timeout parameter and introduces a new state for the main run_cycle loop: starting. When master_start_timeout is zero we will fail over as soon as there is a failover candidate. Otherwise PostgreSQL will be started, but once master_start_timeout expires we will stop and release leader lock if failover is possible. Once failover succeeds or fails (no leader and no one to take the role) we continue with normal processing. While we are waiting for the master timeout we handle manual failover requests.
* Introduce timeout parameter to restart.
When restart timeout is set master becomes eligible for failover after that timeout expires regardless of master_start_time. Immediate restart calls will wait for this timeout to pass, even when node is a standby.
* Fix broken WAL directory symlinks after WAL-E restore.
* Add unit-tests for wale_restore.
* Reduce the amount of MagicMock to the one (for psycopg2.connect)
* Make WAL-E restore process more robuts.
Allow retries only on WAL-E failures.
Sleep after each attempt
* Update the tests.
* Change WAL-E behavior when master is absent, tests.
- Challenge the use of WAL-E even when 'no_master' flag is set. This flag in
fact does not indicate that the master is absent. In order to check the master
absense the script looks whether the connection string is not empty.
- Retry on a failure to fetch current xlog position from the master. The reason
it has to be separate from retries in the main loop is that we don't just
retry the connection attempt, but also make a decision when either it was
successfull or all attempts are exhausted.
- Remove wrong usages of ProperyMocks from the tests.
* Avoid redundant output of the exception message in logger.exception
* Address issues uncovered by flake8
* Add https and auth support for etcd
Also implement support of PATRONI_ETCD_URL and PATRONI_ETCD_SRV
environment variables
* Implement etcd.proxy etcd.cacert, etcd.cert and etcd.key support
Now it should be possible to set up fully encrypted connection to etcd
with authorization.
starting from 9.6 we need wal_level = 'replica' which is alias for 'hot_standby'. It was working before without problems, but if somebody change wal_level to replica, Patroni will expose pending_restart flag, although restart in this case is not necessary.
* bump versions of consul and etcd to the latest for travis integration-tests
If Patroni was started in a docker with pid=1 it will execute itself
with the same arguments. The original process will take care about init
process duties, i.e. handle sigchld and reap dead orphan processes.
Also it will forward SIGINT, SIGHUP, SIGTERM and some other signals to
the real Patroni process.
Previously replicas were always watching for leader key (even if the
postgres was not in the running there). It was not a big issue, but it
was not possible to interrupt such watch in cases if the postgres
started up or stopped successfully. Also it was delaying update_member
call and we had kind of stale information in DCS up to `loop_wait`
seconds. This commit changes such behavior. If the async_executor is
busy by starting/stopping or restarting postgres we will not watch for
leader key but waiting for event from async_executor up to `loop_wait`
seconds. Async executor will fire such event only in case if the
function it was calling returned something what could be evaluated to
boolean True.
Such functionality is really needed to change the way how we are making
decision about necessity of pg_rewind. It will require to have a local
postgres running and for us it is really important to get such
notification as soon as possible.
* Replace pytz.UTC with dateutil.tz.tzutc, it helps to reduce memory by more than 4Mb...
* fix check of python version: 0x0300000 => 0x3000000
* Update leader key before restart and demote
Adds a new configuration variable synchronous_mode. When enabled Patroni will manage synchronous_standby_names to enable synchronous replication whenever there are healthy standbys available. With synchronous mode enabled Patroni will automatically fail over only to a standby that was synchronously replicating at the time of master failure. This effectively means zero lost user visible transactions.
To enforce the synchronous failover guarantee Patroni stores current synchronous replication state in the DCS, using strict ordering, first enable synchronous replication, then publish the information. Standby can use this to verify that it was indeed a synchronous standby before master failed and is allowed to fail over.
We can't enable multiple standbys as synchronous, allowing PostreSQL to pick one because we can't know which one was actually set to be synchronous on the master when it failed. This means that on standby failure commits will be blocked on the master until next run_cycle iteration. TODO: figure out a way to poke Patroni to run sooner or allow for PostgreSQL to pick one without the possibility of lost transactions.
On graceful shutdown standbys will disable themselves by setting a nosync tag for themselves and waiting for the master to notice and pick another standby. This adds a new mechanism for Ha to publish dynamic tags to the DCS.
When the synchronous standby goes away or disconnects a new one is picked and Patroni switches master over to the new one. If no synchronous standby exists Patroni disables synchronous replication (synchronous_standby_names=''), but not synchronous_mode. In this case, only the node that was previously master is allowed to acquire the leader lock.
Added acceptance tests and documentation.
Implementation by @ants with extensive review by @CyberDem0n.
..the same way as for etcd
Change HTTPClient implementation from using `requests.session` to
`urllib3.PoolManager`, because reference implementation from python-consul
didn't really worked with timeouts and was blocking HA loop...
If the Etcd node partitioned from rest of the cluster it is still
possible to read from it (though it returns some stale information),
but it is not possible to write into it.
Previously Patroni was trying to fetch the new cluster view from DCS in
order to figure out is it still the leader or not and Etcd is always
returning stale info where the node still owns the leader key, but with
negative TTL.
This weird bug clearly shows how dangerous premature optimization is.
We are doing number of attempts when trying to initialize replica using
different methods. Any of this attemp may create and put something into
data directory, what causes next attempts fail.
In addition to that improve logging when creating replica.
* reap children before and after running HA loop
When the Patroni is running in a docker container with the pid=1 it is
also responsible for reaping of all dead processes. We can't call
os.waitpid immediately after receiving SIGCHLD because it breaks
subprocess module. It simply stops receiving exit codes of the processes
it executes because these processes. That's why we just registering the
fact of receiving SIGCHLD and reaping children only after execution of
HA loop.
If the postmaster was dying for some reason, Patroni was able to detect
this fact only on the next iteration of HA loop, because zombie
processes where still there and it was possible to send 0 signal to it.
To avoid such situation we should also reap all dead processes before
executing HA loop.
* Don't rely on _cursor_holder when closing connection
it could happen that connection has been opened but not cursor...
* Don't "retry" when fetching current xlog location and it fails
On every iteration of HA loop we are updaing member key in DCS and among
other data there is current xlog location stored in the value.
If the postgres has died for some reason it is not possible to fetch
xlog position and we are just wasting retry_timeout/2 = 5 seconds there.
If this information will be missing from DCS during period of one HA
loop nothing should break. Patroni is not relying on this information
anyway. When it is doing manual or automatic failover it aways
communicates with other nodes directly to get the most fresh
infomation.
* Don't try to update leader optime when postgres is not 100% healthy
`update_lock` method is not only doing update of the leader lock but
also writes the most recent value of xlog position into optime/leader
key. If you know that postgres can be not 100% healthy because it is in
process of restart or recover we should not try to fetch current xlog
position and update 'optime/leader'. Previously we were using
`AsyncExecutor.busy` property for avoiding of such action, but I think
we should be more excpilicit and do the update only if we know that
postgres is 100% healty.
