pt-slave-restart fails on GTID sets with >1 uuid

[Sheeri K. Cabral]

Parallel replication is NOT in use. Using MySQL 5.6.17-log (Oracle version).

On the slave, there's a duplicate key I'd like to skip, so I wanted to use pt-slave-restart.

From the slave:

           Retrieved_Gtid_Set: 70461b4a-1bd5-11e3-b85b-78e3b51c6a18:1471752-1471753,
817a0b2c-74c1-11e4-818a-005056bdf808:382222-382224:398073,
e5499633-1548-11e5-9852-005056bda47b:53-2986
            Executed_Gtid_Set: 70461b4a-1bd5-11e3-b85b-78e3b51c6a18:1471752-1471753,
817a0b2c-74c1-11e4-818a-005056bdf808:382222-382224:398073,
e5499633-1548-11e5-9852-005056bda47b:53-85

[<email address hidden> mysql]# pt-slave-restart
2015-07-24T14:31:05 aus4-2-relay-bin.000013 19297 1062
Argument "53-85" isn't numeric in sort at /usr/bin/pt-slave-restart line 5078.

Seems like it can't handle the Executed_Gtid_Set variable.

[<email address hidden> mysql]# pt-slave-restart --version
pt-slave-restart 2.2.11

Using an up-to-date version.

slave variables:
mysql> show variables like '%slave%';
+------------------------------+-----------------------+
| Variable_name | Value |
+------------------------------+-----------------------+
| init_slave | |
| log_slave_updates | ON |
| log_slow_slave_statements | OFF |
| pseudo_slave_mode | OFF |
| rpl_stop_slave_timeout | 31536000 |
| slave_allow_batching | OFF |
| slave_checkpoint_group | 512 |
| slave_checkpoint_period | 300 |
| slave_compressed_protocol | OFF |
| slave_exec_mode | STRICT |
| slave_load_tmpdir | /tmp |
| slave_max_allowed_packet | 1073741824 |
| slave_net_timeout | 3600 |
| slave_parallel_workers | 0 |
| slave_pending_jobs_size_max | 16777216 |
| slave_rows_search_algorithms | TABLE_SCAN,INDEX_SCAN |
| slave_skip_errors | OFF |
| slave_sql_verify_checksum | ON |
| slave_transaction_retries | 10 |
| slave_type_conversions | |
| sql_slave_skip_counter | 0 |
+------------------------------+-----------------------+
21 rows in set (0.00 sec)

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PT-1295