CONTAP-120359: System Manager becomes unresponsive as MGWD ran out of memory
Issue
- The MGWD process ran out of memory which caused System Manager to become unresponsive.
- Errors were observed in the mgwd.log:
0000002a.3b4b0fa1 0472959f Wed Sep 13 2023 20:22:55 +02:00 [kern_mgwd:info:3153] A [src/rdb/sql_local_unit.cc 806 (0x823d03e00)]: sqlite_logger: SQLite Error: (1) statement aborts at 4: [ATTACH DATABASE '/mroot/etc/cluster_config/rdb/VLDB/_sql/rdb.db' AS vldb;] database vldb is already in use
0000002a.3b4b0fa2 0472959f Wed Sep 13 2023 20:22:55 +02:00 [kern_mgwd:info:3153] 0x823d03e00: 8803eb00002a504a: ERR: SQL_CONTEXT: _execute_setup_stmts:src/sql_context.cc:1056 _execute_setup_stmts: failed to exec statement 'ATTACH DATABASE '/mroot/etc/cluster_config/rdb/VLDB/_sql/rdb.db' AS vldb;' for on demand connection 0x834cabd08: SQL logic error database vldb is already in use
0000002a.3b4b0fa3 0472959f Wed Sep 13 2023 20:22:55 +02:00 [kern_mgwd:info:3153] sqlite exec message: database vldb is already in use
- REST GET requests using the SQL pass-through also fail.