1Z0-888 · Question #68
You are creating a strategy for backing up MySQL using a cold binary backup. The MySQL instance is a replication master with global transaction identifiers (GTIDs) enabled and it uses Transparent…
The correct answer is C. Copy the backup to a remote host. Note: This is a "choose four" question, but the source only marks C - the full correct set is A, C, I, and B, which together satisfy all three requirements. C is correct because the requirement to "allow for a catastrophic hardware failure" means the backup cannot live on the…
Question
- It must be possible to rebuild the instance using the backup.
- It must be encrypted.
- It must allow for a catastrophic hardware failure. Which four steps must be included in the backup strategy? (Choose four.)
Options
- AInclude the keying data and/or configuration in the backup.
- BRestore the backup to a clean MySQL instance.
- CCopy the backup to a remote host.
- DInclude the MySQL socket file in the backup.
- EInclude the ibtmp1 file in the backup.
- FInclude the relay logs in the backup.
- GInclude the operating system disk encryption key in the backup.
- HInclude the MySQL PID file in the backup.
- IInclude the binary logs in the backup.
How the community answered
(37 responses)- B8% (3)
- C65% (24)
- F3% (1)
- H19% (7)
- I5% (2)
Explanation
Note: This is a "choose four" question, but the source only marks C - the full correct set is A, C, I, and B, which together satisfy all three requirements.
C is correct because the requirement to "allow for a catastrophic hardware failure" means the backup cannot live on the same physical host as the data - copying it to a remote host is the only option that survives total hardware loss. A is correct because TDE uses MySQL's keyring plugin to encrypt InnoDB data files; without the keying data/configuration, the backed-up files are permanently unreadable - you cannot rebuild the instance. I is correct because the instance is a GTID-enabled replication master, and binary logs carry the full GTID history; without them, you cannot re-establish replication or perform point-in-time recovery after restore. B is correct because the requirement "it must be possible to rebuild the instance" is only verified by actually restoring to a clean instance - an untested backup cannot be considered valid.
The remaining choices are all distractors: D (socket) and H (PID file) are runtime artifacts MySQL recreates at startup; E (ibtmp1) is a temporary tablespace that holds no persistent data and is also recreated at startup; F (relay logs) belong to replication slaves/replicas, not masters; and G (OS disk encryption key) is irrelevant - TDE is a MySQL-layer feature using its own keyring, entirely separate from OS-level disk encryption.
Memory tip: Runtime files never travel with backups (socket, PID, temp tablespace). For TDE, think "keys travel with the lock" - back up the keyring or the data is forever unreadable. Relay logs are a slave's concern, not a master's. And an unverified backup isn't a backup - always include the restore test.
Topics
Community Discussion
No community discussion yet for this question.