Commit At Most Once v4

The objective of the Commit At Most Once (CAMO) feature is to prevent the application from committing more than once.

Without CAMO, when a client loses connection after a COMMIT is submitted, the application might not receive a reply from the server and is therefore unsure whether the transaction committed.

The application can't easily decide between the two options of:

• Retrying the transaction with the same data, since this can in some cases cause the data to be entered twice

• Not retrying the transaction and risk that the data doesn't get processed at all

Either of those is a critical error with high-value data.

One way to avoid this situation is to make sure that the transaction includes at least one INSERT into a table with a unique index, but that depends on the application design and requires application- specific error-handling logic, so it isn't effective in all cases.

The CAMO feature in BDR offers a more general solution and doesn't require an INSERT. When activated by bdr.enable_camo or bdr.commit_scope, the application receives a message containing the transaction identifier, if already assigned. Otherwise, the first write statement in a transaction sends that information to the client. If the application sends an explicit COMMIT, the protocol ensures that the application receives the notification of the transaction identifier before the COMMIT is sent. If the server doesn't reply to the COMMIT, the application can handle this error by using the transaction identifier to request the final status of the transaction from another BDR node. If the prior transaction status is known, then the application can safely decide whether to retry the transaction.

CAMO works in one of two modes:

• Pair mode
• With Eager All Node Replication

In pair mode, CAMO works by creating a pair of partner nodes that are two BDR master nodes from the same top-level BDR group. In this operation mode, each node in the pair knows the outcome of any recent transaction executed on the other peer and especially (for our need) knows the outcome of any transaction disconnected during COMMIT. The node that receives the transactions from the application might be referred to as "origin" and the node that confirms these transactions as "partner." However, there's no difference in the CAMO configuration for the nodes in the CAMO pair. The pair is symmetric.

When combined with Eager All-Node Replication, CAMO enables every peer (that is, a full BDR master node) to act as a CAMO partner. No designated CAMO partner must be configured in this mode.

Requirements

To use CAMO, an application must issue an explicit COMMIT message as a separate request (not as part of a multi-statement request). CAMO can't provide status for transactions issued from procedures or from single-statement transactions that use implicit commits.

Configuration

Assume an existing EDB Postgres Distributed cluster consists of the nodes node1 and node2. Both nodes are part of a BDR-enabled database called bdrdemo, and both part of the same node group mygroup. You can configure the nodes to be CAMO partners for each other.

1. Create the EDB Postgres Distributed cluster where nodes node1 and node2 are part of the mygroup node group.

2. Run the function bdr.add_camo_pair() on one node:

   → WrapCopy

   SELECT bdr.add_camo_pair('mygroup', 'node1', 'node2');

3. Adjust the application to use the COMMIT error handling that CAMO suggests.

We don't recommend enabling CAMO at the server level, as this imposes higher latency for all transactions, even when not needed. Instead, selectively enable it for individual transactions by turning on CAMO at the session or transaction level.

To enable CAMO at the session level:

SET bdr.enable_camo = 'remote_commit_flush';

To enable CAMO for individual transactions, after starting the transaction and before committing it:

SET LOCAL bdr.enable_camo = 'remote_commit_flush';

Valid values for bdr.enable_camo that enable CAMO are:

• off (default)
• remote_write
• remote_commit_async
• remote_commit_flush or on

See the Comparison of synchronous replication modes for details about how each mode behaves. Setting bdr.enable_camo = off disables this feature, which is the default.

CAMO with Eager All-Node Replication

To use CAMO with Eager All-Node Replication, no changes are required on either node. It is enough to enable the global commit scope after the start of the transaction. You don't need to set bdr.enable_camo.

BEGIN;
SET LOCAL bdr.commit_scope = 'global';
...
COMMIT;

The application still needs to be adjusted to use COMMIT error handling as specified but is free to connect to any available BDR node to query the transaction's status.

Failure scenarios

Different failure scenarios occur in different configurations.

Data persistence at receiver side

By default, a PGL writer operates in bdr.synchronous_commit = off mode when applying transactions from remote nodes. This holds true for CAMO as well, meaning that transactions are confirmed to the origin node possibly before reaching the disk of the CAMO partner. In case of a crash or hardware failure, it is possible for a confirmed transaction to be unrecoverable on the CAMO partner by itself. This isn't an issue as long as the CAMO origin node remains operational, as it redistributes the transaction once the CAMO partner node recovers.

This in turn means CAMO can protect against a single-node failure, which is correct for local mode as well as or even in combination with remote write.

To cover an outage of both nodes of a CAMO pair, you can use bdr.synchronous_commit = local to enforce a flush prior to the pre-commit confirmation. This doesn't work with either remote write or local mode and has a performance impact due to I/O requirements on the CAMO partner in the latency sensitive commit path.

Local mode

When synchronous_replication_availability = 'async', a node (i.e., master) detects whether its CAMO partner is ready. If not, it temporarily switches to local mode. When in local mode, a node commits transactions locally until switching back to CAMO mode.

This doesn't allow COMMIT status to be retrieved, but it does let you choose availability over consistency. This mode can tolerate a single-node failure. In case both nodes of a CAMO pair fail, they might choose incongruent commit decisions to maintain availability, leading to data inconsistencies.

For a CAMO partner to switch to ready, it needs to be connected, and the estimated catchup interval needs to drop below bdr.global_commit_timeout. The current readiness status of a CAMO partner can be checked with bdr.is_camo_partner_ready, while bdr.node_replication_rates provides the current estimate of the catchup time.

The switch from CAMO protected to local mode is only ever triggered by an actual CAMO transaction either because the commit exceeds the bdr.global_commit_timeout or, in case the CAMO partner is already known, disconnected at the time of commit. This switch is independent of the estimated catchup interval. If the CAMO pair is configured to require Raft to switch to local mode, this switch requires a majority of nodes to be operational (see the require_raft flag for bdr.add_camo_pair). This can prevent a split brain situation due to an isolated node from switching to local mode. If require_raft isn't set for the CAMO pair, the origin node switches to local mode immediately.

You can configure the detection on the sending node using PostgreSQL settings controlling keep-alives and timeouts on the TCP connection to the CAMO partner. The wal_sender_timeout is the time that a node waits for a CAMO partner until switching to local mode. Additionally, the bdr.global_commit_timeout setting puts a per-transaction limit on the maximum delay a COMMIT can incur due to the CAMO partner being unreachable. It might be lower than the wal_sender_timeout, which influences synchronous standbys as well, and for which a good compromise between responsiveness and stability must be found.

The switch from local mode to CAMO mode depends on the CAMO partner node, which initiates the connection. The CAMO partner tries to reconnect at least every 30 seconds. After connectivity is reestablished, it might therefore take up to 30 seconds until the CAMO partner connects back to its origin node. Any lag that accumulated on the CAMO partner further delays the switch back to CAMO protected mode.

Unlike during normal CAMO operation, in local mode there's no additional commit overhead. This can be problematic, as it allows the node to continuously process more transactions than the CAMO pair can normally process. Even if the CAMO partner eventually reconnects and applies transactions, its lag only ever increases in such a situation, preventing reestablishing the CAMO protection. To artificially throttle transactional throughput, BDR provides the bdr.camo_local_mode_delay setting, which allows you to delay a COMMIT in local mode by an arbitrary amount of time. We recommend measuring commit times in normal CAMO mode during expected workloads and configuring this delay accordingly. The default is 5 ms, which reflects a local network and a relatively quick CAMO partner response.

Consider the choice of whether to allow local mode in view of the architecture and the availability requirements. The following examples provide some detail.

Example: Symmetric node pair

This example considers a setup with two BDR nodes that are the CAMO partner of each other. This is the only possible configuration starting with BDR4.

This configuration enables CAMO behavior on both nodes. It's therefore suitable for workload patterns where it is acceptable to write concurrently on more than one node, such as in cases that aren't likely to generate conflicts.

With local mode

If local mode is allowed, there's no single point of failure. When one node fails:

• The other node can determine the status of all transactions that were disconnected during COMMIT on the failed node.
• New write transactions are allowed:
  • If the second node also fails, then the outcome of those transactions that were being committed at that time is unknown.

Without local mode

If local mode isn't allowed, then each node requires the other node for committing transactions, that is, each node is a single point of failure. When one node fails:

• The other node can determine the status of all transactions that were disconnected during COMMIT on the failed node.
• New write transactions are prevented until the node recovers.

Application use

Overview and requirements

CAMO relies on a retry loop and specific error handling on the client side. There are three aspects to it:

• The result of a transaction's COMMIT needs to be checked and, in case of a temporary error, the client must retry the transaction.
• Prior to COMMIT, the client must retrieve a global identifier for the transaction, consisting of a node id and a transaction id (both 32-bit integers).
• If the current server fails while attempting a COMMIT of a transaction, the application must connect to its CAMO partner, retrieve the status of that transaction, and retry depending on the response.

The application must store the global transaction identifier only for the purpose of verifying the transaction status in case of disconnection during COMMIT. In particular, the application doesn't need an additional persistence layer. If the application fails, it needs only the information in the database to restart.

Adding a CAMO pair

The function bdr.add_camo_pair() configures an existing pair of BDR nodes to work as a symmetric CAMO pair.

The require_raft option controls how and when to switch to local mode in case synchronous_replication_availability is set to async, allowing such a switch in general.

Synopsis

bdr.add_camo_pair(node_group text, left_node text, right_node text,
                  require_raft boolean)

Changing the configuration of a CAMO pair

The function bdr.alter_camo_pair() allows you to toggle the require_raft You can't currently change the nodes of a pairing. You must instead use bdr.remove_camo_pair followed by bdr.add_camo_pair.

Synopsis

bdr.alter_camo_pair(node_group text, left_node text, right_node text,
                    require_raft boolean)

Removing a CAMO pair

The function bdr.remove_camo_pair() removes a CAMO pairing of two nodes and disallows future use of CAMO transactions by bdr.enable_camo on those two nodes.

Synopsis

bdr.remove_camo_pair(node_group text, left_node text, right_node text)

CAMO partner connection status

The function bdr.is_camo_partner_connected allows checking the connection status of a CAMO partner node configured in pair mode. There currently is no equivalent for CAMO used with Eager Replication.

Synopsis

bdr.is_camo_partner_connected()

Return value

A Boolean value indicating whether the CAMO partner is currently connected to a WAL sender process on the local node and therefore can receive transactional data and send back confirmations.

CAMO partner readiness

The function bdr.is_camo_partner_ready allows checking the readiness status of a CAMO partner node configured in pair mode. Underneath, this triggers the switch to and from local mode.

Synopsis

bdr.is_camo_partner_ready()

Return value

A Boolean value indicating whether the CAMO partner can reasonably be expected to confirm transactions originating from the local node in a timely manner (before bdr.global_commit_timeout expires).

Fetch the CAMO partner

This function shows the local node's CAMO partner (configured by pair mode).

bdr.get_configured_camo_partner()

Wait for consumption of the apply queue from the CAMO partner

The function bdr.wait_for_camo_partner_queue is a wrapper of bdr.wait_for_apply_queue defaulting to query the CAMO partner node. It yields an error if the local node isn't part of a CAMO pair.

Synopsis

bdr.wait_for_camo_partner_queue()

Transaction status between CAMO nodes

This function enables a wait for CAMO transactions to be fully resolved.

bdr.camo_transactions_resolved()

Transaction status query function

To check the status of a transaction that was being committed when the node failed, the application must use this function:

bdr.logical_transaction_status(node_id OID, xid OID,
                               require_camo_partner boolean)

With CAMO used in pair mode, use this function only on a node that's part of a CAMO pair. Along with Eager replication, you can use it on all nodes.

In both cases, you must call the function within 15 minutes after the commit was issued. The CAMO partner must regularly purge such meta-information and therefore can't provide correct answers for older transactions.

Before querying the status of a transaction, this function waits for the receive queue to be consumed and fully applied. This prevents early negative answers for transactions that were received but not yet applied.

Despite its name, it's not always a read-only operation. If the status is unknown, the CAMO partner decides whether to commit or abort the transaction, storing that decision locally to ensure consistency going forward.

The client must not call this function before attempting to commit on the origin. Otherwise the transaction might be forced to roll back.

Synopsis

bdr.logical_transaction_status(node_id OID, xid OID,
                               require_camo_partner boolean DEFAULT true)

Parameters

• node_id — The node id of the BDR node the transaction originates from, usually retrieved by the client before COMMIT from the PQ parameter bdr.local_node_id.
• xid — The transaction id on the origin node, usually retrieved by the client before COMMIT from the PQ parameter transaction_id (requires enable_camo to be set to on, remote_write, remote_commit_async, or remote_commit_flush. See Commit At Most Once settings)
• require_camo_partner — Defaults to true and enables configuration checks. Set to false to disable these checks and query the status of a transaction that was protected by Eager All-Node Replication.

Return value

The function returns one of these results:

• 'committed'::TEXT — The transaction was committed, is visible on both nodes of the CAMO pair, and will eventually be replicated to all other BDR nodes. No need for the client to retry it.

• 'aborted'::TEXT — The transaction was aborted and will not be replicated to any other BDR node. The client needs to either retry it or escalate the failure to commit the transaction.

• 'in progress'::TEXT — The transaction is still in progress on this local node and wasn't committed or aborted yet. The transaction might be in the COMMIT phase, waiting for the CAMO partner to confirm or deny the commit. The recommended client reaction is to disconnect from the origin node and reconnect to the CAMO partner to query that instead. With a load balancer or proxy in between, where the client lacks control over which node gets queried, the client can only poll repeatedly until the status switches to either 'committed' or 'aborted'.

  For Eager All-Node Replication, peer nodes yield this result for transactions that aren't yet committed or aborted. This means that even transactions not yet replicated (or not even started on the origin node) might yield an in progress result on a peer BDR node in this case. However, the client must not query the transaction status prior to attempting to commit on the origin.

• 'unknown'::TEXT — The transaction specified is unknown, either because it's in the future, not replicated to that specific node yet, or too far in the past. The status of such a transaction is not yet or no longer known. This return value is a sign of improper use by the client.

The client must be prepared to retry the function call on error.

Connection pools and proxies

The effect of connection pools and proxies needs to be considered when designing a CAMO cluster. A proxy may freely distribute transactions to all nodes in the commit group (i.e. to both nodes of a CAMO pair or to all BDR nodes in case of Eager All Node Replication).

Care needs to be taken to ensure that the application fetches the proper node id: when using session pooling, the client remains connected to the same node, so the node id remains constant for the lifetime of the client session. However, with finer-grained transaction pooling, the client needs to fetch the node id for every transaction (as in the example given below).

A client that is not directly connected to the BDR nodes might not even notice a failover or switchover, but can always use the bdr.local_node_id parameter to determine which node it is currently connected to. In the crucial situation of a disconnect during COMMIT, the proxy must properly forward that disconnect as an error to the client applying the CAMO protocol.

For CAMO in remote_write mode, a proxy that potentially switches between the CAMO pairs must use the bdr.wait_for_camo_partner_queue function to prevent stale reads.

HARP is the only proxy that supports all of the above requirements. PgBouncer and HAproxy can work with CAMO, but do not support CAMO's remote_write mode.

Example

The following example demonstrates what a retry loop of a CAMO aware client application should look like in C-like pseudo-code. It expects two DSNs origin_dsn and partner_dsn providing connection information. These usually are the same DSNs as used for the initial call to bdr.create_node, and can be looked up in bdr.node_summary, column interface_connstr.

PGconn *conn = PQconnectdb(origin_dsn);

loop {
    // start a transaction
    PQexec(conn, "BEGIN");

    // apply transactional changes
    PQexec(conn, "INSERT INTO ...");
    ...

    // store a globally unique transaction identifier
    node_id = PQparameterStatus(conn, "bdr.local_node_id");
    xid = PQparameterStatus(conn, "transaction_id");

    // attempt to commit
    PQexec(conn, "COMMIT");
    if (PQresultStatus(res) == PGRES_COMMAND_OK)
        return SUCCESS;
    else if (PQstatus(res) == CONNECTION_BAD)
    {
        // Re-connect to the partner
        conn = PQconnectdb(partner_dsn);
        // Check if successfully reconnected
        if (!connectionEstablished())
            panic();

        // Check the attempted transaction's status
        sql = "SELECT bdr.logical_transaction_status($node_id, $xid)";
        txn_status = PQexec(conn, sql);
        if (txn_status == "committed")
            return SUCCESS;
        else
            continue;   // to retry the transaction on the partner
    }
    else
    {
        // The connection is intact, but the transaction failed for some
        // other reason.  Differentiate between permanent and temporary
        // errors.
        if (isPermanentError())
            return FAILURE;
        else
        {
            // Determine an appropriate delay to back-off to account for
            // temporary failures due to congestion, so as to decrease
            // the overall load put on the servers.
            sleep(increasing_retry_delay);

            continue;
        }
    }
}

This example needs to be extended with proper logic for connecting, including retries and error handling. If using a load balancer (e.g. PgBouncer), re-connecting can be implemented by simply using PQreset. Ensure that the load balancer only ever redirects a client to a CAMO partner and not any other BDR node.

In practice, an upper limit of retries is recommended. Depending on the actions performed in the transaction, other temporary errors may be possible and need to be handled by retrying the transaction depending on the error code, similarly to the best practices on deadlocks or on serialization failures while in SERIALIZABLE isolation mode.

Interaction with DDL and global locks

Transactions protected by CAMO can contain DDL operations. However, DDL uses global locks, which already provide some synchronization among nodes. See DDL locking details for more information.

Combining CAMO with DDL imposes a higher latency and also increases the chance of global deadlocks. We therefore recommend using a relatively low bdr.global_lock_timeout, which aborts the DDL and therefore resolves a deadlock in a reasonable amount of time.

Nontransactional DDL

The following DDL operations aren't allowed in a transaction block and therefore can't benefit from CAMO protection. For these, CAMO is automatically disabled internally:

• all concurrent index operations (CREATE, DROP, and REINDEX)
• REINDEX DATABASE, REINDEX SCHEMA, and REINDEX SYSTEM
• VACUUM
• CLUSTER without any parameter
• ALTER TABLE DETACH PARTITION CONCURRENTLY
• ALTER TYPE [enum] ADD VALUE
• ALTER SYSTEM
• CREATE and DROP DATABASE
• CREATE and DROP TABLESPACE
• ALTER DATABASE [db] TABLESPACE

CAMO limitations

• CAMO is designed to query the results of a recently failed COMMIT on the origin node, so in case of disconnection, code the application to immediately request the transaction status from the CAMO partner. Have as little delay as possible after the failure before requesting the status. Applications must not rely on CAMO decisions being stored for longer than 15 minutes.

• If the application forgets the global identifier assigned, for example as a result of a restart, there's no easy way to recover it. Therefore, we recommend that applications wait for outstanding transactions to end before shutting down.

• For the client to apply proper checks, a transaction protected by CAMO can't be a single statement with implicit transaction control. You also can't use CAMO with a transaction-controlling procedure or in a DO block that tries to start or end transactions.

• CAMO resolves commit status but doesn't yet resolve pending notifications on commit. CAMO and Eager replication options don't allow the NOTIFY SQL command or the pg_notify() function. They also don't allow LISTEN or UNLISTEN.

• When replaying changes, CAMO transactions may detect conflicts just the same as other transactions. If timestamp conflict detection is used, the CAMO transaction uses the timestamp of the prepare on the origin node, which is before the transaction becomes visible on the origin node itself.

• CAMO is not currently compatible with transaction streaming. Please ensure to disable transaction streaming when planning to use CAMO. This can be configured globally or in the BDR node group, see Transaction Streaming Configuration.

Performance implications

CAMO extends the Postgres replication protocol by adding a message roundtrip at commit. Applications have a higher commit latency than with asynchronous replication, mostly determined by the roundtrip time between involved nodes. Increasing the number of concurrent sessions can help to increase parallelism to obtain reasonable transaction throughput.

The CAMO partner confirming transactions must store transaction states. Compared to non-CAMO operation, this might require an additional seek for each transaction applied from the origin.

Client application testing

Proper use of CAMO on the client side isn't trivial. We strongly recommend testing the application behavior with the BDR cluster against failure scenarios such as node crashes or network outages.

CAMO versus group commit

CAMO doesn't currently work with group commit.