influxdata/influxdb · error

no duplicate table by ID or name

Error message

no duplicate table by ID or name

What it means

After building a TableTransaction for a newly created table, the code inserts it into self.tables and expects no entry with the same table_id (or name-keyed duplicate) to exist. A hit means the batch already contains a table with this ID or name, violating the one-transaction-per-table invariant for new-table creation paths.

Solutions

1. Ensure each table is created at most once per CatalogBatch; deduplicate table creation before building the batch
2. Never replay or append to a batch that has already had a table inserted; start a fresh batch on retry
3. Confirm table_id generation is unique (no reset/colliding sequence) in any custom batch construction
4. If triggered through the normal write API on a current version, report it as a bug

Example fix

// before
batch.create_table("cpu")?;
batch.create_table("cpu")?; // panics on insert: duplicate by ID/name
// after
if !batch.contains_table("cpu") {
    batch.create_table("cpu")?;
}

// dedupe table creations before applying
let mut seen = std::collections::HashSet::new();
for name in &table_names {
    if !seen.insert(name.clone()) {
        return Err(anyhow!("duplicate table creation for {name} in one batch"));
    }
}

fn is_new_table(batch: &CatalogBatch, id: TableId, name: &str) -> bool {
    !batch.tables.keys().any(|k| k.id == id)
        && batch.tables.values().all(|t| t.table.name != name)
}

When it happens

Trigger: Calling the create-table path (which ends in this insert) with a table_id/name that already has a transaction in the batch — e.g. creating the same table twice in one CatalogBatch, or a table_id collision from a mis-seeded generator.

Common situations: Duplicate CREATE TABLE statements consolidated into a single batch; a batch replayed/appended after partial application; custom tooling reusing table IDs across batches.

Understand the failure class

Background: "This is a bug, please report it": internal invariant violations, unreachable panics, and SNH errors explained — this error's family across 47 libraries.

Related errors

• column id in series key should be valid
• Existing transaction for table should not exist
• ordered catalog batch should succeed when applied
• table should exist by id
• attempted to create a resource that already exists

AI-assisted analysis of influxdata/influxdb@06200ef96b (2026-09-19). Data as JSON: /api/errors/cc995cb92d7a16cf. Report an issue: GitHub.

Appendix: source

Thrown at influxdb3_catalog/src/catalog/versions/v2/update.rs:2030

            self.columns_per_table_limit,
            self.storage_mode,
        );

        if let Some(CreateTableColumns { tags, fields }) = columns {
            for tag in tags {
                table_tx.add_tag(tag)?;
            }

            for (field_name, field_type) in fields {
                table_tx.add_field(field_name, (*field_type).into())?;
            }

            table_tx.add_time()?;
        }

        self.tables
            .insert(table_id, table_tx)
            .expect("no duplicate table by ID or name");

        Ok(table_id)
    }

    pub fn db_schema(&self) -> &Arc<DatabaseSchema> {
        &self.database_schema
    }
}

impl From<DatabaseCatalogTransaction> for CatalogBatch {
    fn from(txn: DatabaseCatalogTransaction) -> Self {
        let mut ops = txn.ops;
        let mut tables = txn.tables;
        ops.extend(tables.repo.drain(..).filter_map(|(_, tx)| {
            if tx.column_definitions.is_empty() && tx.field_family_definitions.is_empty() {
                None
            } else {
                Some(DatabaseCatalogOp::AddColumns(tx.into()))

View on GitHub (pinned to 06200ef96b)