What causes "Cannot mix IF conditions and %s for the same row" in apache/cassandra?

Use only one kind of condition per row: either column conditions (IF col=val) or existence conditions (IF EXISTS / IF NOT EXISTS), never both for the same primary key.

apache/cassandra · error · InvalidRequestException

Cannot mix IF conditions and %s for the same row

Error message

Cannot mix IF conditions and %s for the same row

What it means

Cassandra's lightweight-transaction (CAS) layer requires that all conditions for a given row be of the same kind. CQL3CasRequest.addConditions throws InvalidRequestException when a second batch of conditions is added for a row whose existing condition is not a ColumnsConditions (i.e., the row already has an EXISTS/NOT_EXISTS condition or a different condition type), because mixing IF column conditions and row-existence checks on the same row cannot be evaluated as one atomic predicate.

Source

Thrown at src/java/org/apache/cassandra/cql3/statements/CQL3CasRequest.java:209

                    : new InvalidRequestException("Cannot mix IF conditions and IF " + (isNotExist ? "NOT " : "") + "EXISTS for the same row");
            }
        }

        setConditionsForRow(clustering, condition);
        hasExists = true;
    }

    public void addConditions(Clustering<?> clustering, Collection<ColumnCondition> conds, QueryOptions options) throws InvalidRequestException
    {
        RowCondition condition = getConditionsForRow(clustering);
        if (condition == null)
        {
            condition = new ColumnsConditions(clustering);
            setConditionsForRow(clustering, condition);
        }
        else if (!(condition instanceof ColumnsConditions))
        {
            throw new InvalidRequestException("Cannot mix IF conditions and " + ((ToCQL) condition).toCQL() + " for the same row");
        }
        ((ColumnsConditions)condition).addConditions(conds, options);
    }

    private RowCondition getConditionsForRow(Clustering<?> clustering)
    {
        return clustering == Clustering.STATIC_CLUSTERING ? staticConditions : conditions.get(clustering);
    }

    private void setConditionsForRow(Clustering<?> clustering, RowCondition condition)
    {
        if (clustering == Clustering.STATIC_CLUSTERING)
        {
            assert staticConditions == null;
            staticConditions = condition;
        }
        else
        {

View on GitHub (pinned to 88fd0f6a0e)

Solutions

Use only one kind of condition per row: either column conditions (IF col=val) or existence conditions (IF EXISTS / IF NOT EXISTS), never both for the same primary key.
If both semantics are needed, split into two statements on different rows or use a single column condition that implies existence (e.g., IF somecol = expected).
In custom code calling CQL3CasRequest, check getConditionsForRow(clustering) instanceof ColumnsConditions before adding more conditions, or create a fresh CQL3CasRequest.

Example fix

// before
UPDATE ks.t SET v=1 WHERE k=1 IF NOT EXISTS; -- combined with
UPDATE ks.t SET v=2 WHERE k=1 IF v=0;       -- same row, mixed kinds
// after
UPDATE ks.t SET v=1 WHERE k=1 IF v=0; -- single column condition only

Defensive patterns

Strategy: validation

Validate before calling

// CQL-level guard: never combine existence and column conditions on the same PK row
boolean hasExists = cql.matches("(?s).*IF\\s+(NOT\\s+)?EXISTS.*");
boolean hasColumnCond = cql.matches("(?s).*IF\\s+[a-zA-Z_][a-zA-Z0-9_]*\\s*(=|<|>|!=|IN|CONTAINS).*");
if (hasExists && hasColumnCond) throw new IllegalArgumentException("Mix IF EXISTS and column conditions only on different rows");

Try / catch

try { session.execute(cql); } catch (InvalidRequestException e) { if (e.getMessage().contains("Cannot mix IF conditions")) { /* split statement or drop one condition kind */ } else throw e; }

Prevention

Keep at most one condition kind (existence or column comparison) per primary-key row per CAS request
When batching LWT statements, ensure conditional statements target distinct rows
Review driver LWT builder code for repeated addConditions calls on the same clustering key

When it happens

Trigger: Executing an INSERT/UPDATE/DELETE with LWT conditions that combine a plain existence check (IF EXISTS / IF NOT EXISTS) with column-value conditions (e.g., IF col = x) targeting the same primary-key row in one batch or statement flow, so addConditionsTo invokes addConditions twice for the same clustering with mismatched condition types.

Common situations: Hand-building CAS requests in internal code or drivers' LWT builders; batching a conditional INSERT ... IF NOT EXISTS together with UPDATE ... IF col=val on the same row; refactoring statement code that previously applied conditions to different rows.

Understand the failure class

Background: Conflicting config options: "cannot be used together" — configuration validation errors across open-source libraries — this error's family across 162 libraries.

Related errors

AI-assisted analysis of apache/cassandra@88fd0f6a0e (2026-09-10). Data as JSON: /api/errors/f2909ba76f4d1a20. Report an issue: GitHub.