prestodb/presto · error · IllegalStateException

Cannot perform Metastore updates in replay mode

Error message

Cannot perform Metastore updates in replay mode

What it means

verifyRecordingMode is a shared guard used by all metastore update operations (create/drop/rename, statistics updates, etc.): in replay mode the metastore is read-only against a recorded log, so any write/update request is rejected with this IllegalStateException to protect the recording's integrity.

Source

Thrown at presto-hive-metastore/src/main/java/com/facebook/presto/hive/metastore/RecordingHiveMetastore.java:538

        return delegate.addConstraint(metastoreContext, databaseName, tableName, tableConstraint);
    }

    private <K, V> V loadValue(Cache<K, V> cache, K key, Supplier<V> valueSupplier)
    {
        if (replay) {
            return Optional.ofNullable(cache.getIfPresent(key))
                    .orElseThrow(() -> new PrestoException(NOT_FOUND, "Missing entry found for key: " + key));
        }

        V value = valueSupplier.get();
        cache.put(key, value);
        return value;
    }

    private void verifyRecordingMode()
    {
        if (replay) {
            throw new IllegalStateException("Cannot perform Metastore updates in replay mode");
        }
    }

    @Immutable
    public static class Recording
    {
        private final Optional<List<String>> allDatabases;
        private final Optional<Set<String>> allRoles;
        private final List<Pair<String, Optional<Database>>> databases;
        private final List<Pair<HiveTableHandle, Optional<Table>>> tables;
        private final List<Pair<HiveTableName, List<TableConstraint<String>>>> tableConstraints;
        private final List<Pair<String, Set<ColumnStatisticType>>> supportedColumnStatistics;
        private final List<Pair<HiveTableName, PartitionStatistics>> tableStatistics;
        private final List<Pair<Set<HivePartitionName>, Map<String, PartitionStatistics>>> partitionStatistics;
        private final List<Pair<String, Optional<List<String>>>> allTables;
        private final List<Pair<String, Optional<List<String>>>> allViews;
        private final List<Pair<HivePartitionName, Optional<Partition>>> partitions;
        private final List<Pair<HiveTableName, Optional<List<PartitionNameWithVersion>>>> partitionNames;

View on GitHub (pinned to 55bb57d202)

Solutions

1. Run the cluster in recording mode (replay disabled) if write operations are expected
2. Replay sessions should only issue read operations; move updates to a non-replay deployment

When it happens

Trigger: Thrown at presto-hive-metastore/src/main/java/com/facebook/presto/hive/metastore/RecordingHiveMetastore.java:538 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of prestodb/presto@55bb57d202 (2026-09-04). Data as JSON: /api/errors/0237bf5d77dbee01. Report an issue: GitHub.