replication show slots v5.6
Shows the status of BDR replication slots
Synopsis
Shows the status of BDR replication slots. Output with the verbose flag gives details such as is slot active, replication state (disconnected, streaming, catchup), and approximate lag.
| Symbol | Meaning |
|---|---|
* | ok |
~ | warning (lag > 10M) |
! | critical (lag > 100M OR slot is 'inactive' OR 'disconnected') |
x | down / unreachable |
- | n/a |
In matrix view, sometimes byte lag is shown in parentheses. It is maxOf(WriteLag, FlushLag, ReplayLag, SentLag).
pgd replication show slots [flags]
Options
| Flag | Flag, long | Value | Description |
|---|---|---|---|
--verbose | Verbose mode - provides mode detailed information |
Examples
In a cluster with a node down
In this example, there is a 3 node cluster, bdr-a1 and bdr-c1 are up, bdr-b1 is down.
$ pgd replication show slots
Output
Node bdr-a1 bdr-b1 bdr-c1 ---- ------ ------ ------ bdr-a1 * !(6.6G) * bdr-b1 x * x bdr-c1 * !(6.9G) *
Or you can use the verbose flag to get more details.
$ pgd replication show slots --verboseOutput
Origin Node Target Node Status (active/state) Write Lag (bytes/duration) Flush Lag (bytes/duration) Replay Lag (bytes/duration) Sent Lag (bytes) ----------- ----------- --------------------- -------------------------- -------------------------- --------------------------- ---------------- bdr-a1 bdr-b1 f / disconnected 6.6G / 8 days 02:58:36.243723 6.6G / 8 days 02:58:36.243723 6.6G / 8 days 02:58:36.243723 6.6G bdr-a1 bdr-c1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-c1 bdr-a1 t / streaming 0B / 00:00:00.000812 0B / 00:00:00.000812 0B / 00:00:00.000812 0B bdr-c1 bdr-b1 f / disconnected 6.9G / 8 days 02:58:36.004415 6.9G / 8 days 02:58:36.004415 6.9G / 8 days 02:58:36.004415 6.9G
In a restarted cluster
In this example, there is a 3 node cluster. Node bdr-b1 was down and it has just been restarted.
$ pgd replication show slots
Output
Node bdr-a1 bdr-b1 bdr-c1 ---- ------ ------ ------ bdr-a1 * !(6.9G) * bdr-b1 * * * bdr-c1 * !(5.8G) *
The verbose flag gives more details.
$ pgd replication show slots --verboseOutput
Origin Node Target Node Status (active/state) Write Lag (bytes/duration) Flush Lag (bytes/duration) Replay Lag (bytes/duration) Sent Lag (bytes) ----------- ----------- --------------------- -------------------------- -------------------------- --------------------------- ---------------- bdr-a1 bdr-b1 t / catchup 6.9G / 00:00:00.000778 6.9G / 00:00:00.000778 6.9G / 00:00:00.000778 6.9G bdr-a1 bdr-c1 t / streaming 0B / 00:00:00.104121 0B / 00:00:00.104133 0B / 00:00:00.104133 0B bdr-b1 bdr-a1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-b1 bdr-c1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-c1 bdr-a1 t / streaming 6.8K / 00:00:00 6.8K / 00:00:00 6.8K / 00:00:00 6.8K bdr-c1 bdr-b1 t / catchup 5.5G / 00:00:00.008257 5.5G / 00:00:00.008257 5.5G / 00:00:00.008257 5.5G
In a streaming cluster
In this example, there is a 3 node cluster, all nodes are up and in 'streaming' state.
$ pgd replication show slots
Output
Node bdr-a1 bdr-b1 bdr-c1 ---- ------ ------ ------ bdr-a1 * * * bdr-b1 * * * bdr-c1 * * *
The verbose flag gives more details.
$ pgd replication show slots --verboseOutput
Origin Node Target Node Status (active/state) Write Lag (bytes/duration) Flush Lag (bytes/duration) Replay Lag (bytes/duration) Sent Lag (bytes) ----------- ----------- --------------------- -------------------------- -------------------------- --------------------------- ---------------- bdr-a1 bdr-b1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-a1 bdr-c1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-b1 bdr-a1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-b1 bdr-c1 t / streaming 0B / 00:00:00 0B / 00:00:00 0B / 00:00:00 0B bdr-c1 bdr-a1 t / streaming 0B / 00:00:00 528B / 00:00:00 528B / 00:00:00 0B bdr-c1 bdr-b1 t / streaming 528B / 00:00:00 528B / 00:00:00 528B / 00:00:00 0B
With other node types
In this example the cluster has witness, logical standby and subscriber-only nodes. Upstream for logical-standby-a1 is bdr-a1 and logical-standby-b1 is bdr-b1.
- A logical standby is sent data only by one source node, but no other nodes receive replication changes from it.
- A Subscriber-only node subscribes to replication changes from other nodes in the cluster, but no other nodes receive replication changes from it.
$ pgd replication show slots
Output
Node bdr-a1 bdr-b1 logical-standby-a1 logical-standby-b1 subscriber-only-c1 witness-c1 ---- ------ ------ ------------------ ------------------ ------------------ ---------- bdr-a1 * * * - * * bdr-b1 * * - * * * logical-standby-a1 - - * - - - logical-standby-b1 - - - * - - subscriber-only-c1 - - - - * - witness-c1 * * - - * *