AWS Database Migration Service (DMS) - SHOW BINARY LOGS

Question

Our current setup has a AWS DMS replication instance where 1 migration task is performing CDC reading from a source endpoint (Aurora MySQL writer instance) to a target endpoint (Redshift), after a full load.

I'm getting pinged by my Aurora MySQL database administrator that we're seeing "SHOW BINARY LOG" calls like what we have below 20-50 times a day, coming from DMS (private IP address matches that of the replication instance).

```
@logStreamaurora-mysql-[redacted]
@message# Time: 2023-09-13T07:41:16.230048Z
# User@Host: [redacted][redacted] @  [redacted] Id: [redacted]
# Query_time: 27.554365  Lock_time: 0.000000 Rows_sent: 0  Rows_examined: 0
SET timestamp=1694590848;
SHOW BINARY LOGS;
```

Any idea why AWS DMS is executing this? And if it adds to the CPU load of our Aurora MySQL cluster?

Answer

Hi,

This stack exchange question https://dba.stackexchange.com/questions/316433/how-to-speed-up-show-binary-logs will be very useful to you.

Some folks have noticed same behavior re SHOW BINARY LOG and propose solutions.

Best,

Didier

Answer

Hi Didier,

Thank you for your response. If I understand the stack exchange information correctly, each DMS task runs every few seconds "SHOW BINARY LOGS" command to get the list of binlogs to read from, for the purposes of CDC.

Is this a configurable schedule in DMS? And how would it affect CDC latency/performance overall?

Best,
Zyntas

AWS Database Migration Service (DMS) - SHOW BINARY LOGS

Contenuto pertinente