What causes "Error reading partition %s (stacktrace follows):" in apache/cassandra?

Read the logged stacktrace to identify the corruption type (EOF vs CRC vs deserialize)

Error reading partition %s (stacktrace follows):

Error message

Error reading partition %s (stacktrace follows):

What it means

The scrubber caught a non-fatal Throwable while deserializing a partition from the data file. It logs the error with stacktrace via the output handler, then (when index data is available) attempts to retry the partition from the index-recorded position instead of failing the whole scrub.

Source

Thrown at src/java/org/apache/cassandra/io/sstable/format/big/BigTableScrubber.java:175

                {
                    throw new IOError(new IOException(String.format("Key from data file (%s) does not match key from index file (%s)",
                                                                    //ByteBufferUtil.bytesToHex(key.getKey()), ByteBufferUtil.bytesToHex(currentIndexKey))));
                                                                    "_too big_", ByteBufferUtil.bytesToHex(currentIndexKey))));
                }

                if (indexFile != null && dataSizeFromIndex > dataFile.length())
                    throw new IOError(new IOException("Impossible partition size (greater than file length): " + dataSizeFromIndex));

                if (indexFile != null && dataStart != dataStartFromIndex)
                    outputHandler.warn("Data file partition position %d differs from index file row position %d", dataStart, dataStartFromIndex);

                if (tryAppend(prevKey, key, writer))
                    prevKey = key;
            }
            catch (Throwable th)
            {
                throwIfFatal(th);
                outputHandler.warn(th, "Error reading partition %s (stacktrace follows):", keyName);

                if (currentIndexKey != null
                    && (key == null || !key.getKey().equals(currentIndexKey) || dataStart != dataStartFromIndex))
                {

                    outputHandler.output("Retrying from partition index; data is %s bytes starting at %s",
                                         dataSizeFromIndex, dataStartFromIndex);
                    key = sstable.decorateKey(currentIndexKey);
                    try
                    {
                        if (!tableMetadata.isIndex())
                            tableMetadata.partitionKeyType.validate(key.getKey());
                        dataFile.seek(dataStartFromIndex);

                        if (tryAppend(prevKey, key, writer))
                            prevKey = key;
                    }
                    catch (Throwable th2)

View on GitHub (pinned to 88fd0f6a0e)

Solutions

Read the logged stacktrace to identify the corruption type (EOF vs CRC vs deserialize)
Re-run scrub letting it skip bad partitions, then run `nodetool repair` to restore lost data
Restore the affected sstable from a snapshot or another replica
Run `nodetool verify` to scope the corruption and check compression checksums

Example fix

// before: scrubbing a corrupt partition throws and warns
// after: run verify first to scope damage
nodetool verify keyspace table;
nodetool scrub keyspace table;
nodetool repair keyspace table;

Defensive patterns

Strategy: try-catch

Try / catch

try {
    scrubPartition(dataFile);
} catch (CorruptSSTableException | IOException e) {
    logger.warn("Skipping corrupt partition at {} in {}", dataFile.getFilePointer(), sstable, e);
    skipToNextPartition();
}

Prevention

Run `nodetool verify` regularly to catch corruption early
Use checksummed sstable formats (modern versions checksum compression chunks)
Keep replicas (RF >= 3) so repair can restore skipped partitions
Enable filesystem checksums to detect bit-rot

When it happens

Trigger: scrubInternal throws while reading a partition (CorruptSSTableException, EOFException, IOError) — corrupted bytes in the -Data.db at the current partition, or a cell deserialization error.

Common situations: Bit-rot on disk, truncated data files after crashes, corrupt compression chunks, or cells written by an incompatible format version.

Understand the failure class

Background: "failed to read file", EACCES, ENOENT and "could not read <path>" errors: when a program can't read a file from disk — this error's family across 49 libraries.

Related errors

AI-assisted analysis of apache/cassandra@88fd0f6a0e (2026-09-10). Data as JSON: /api/errors/7b2fa707e334f067. Report an issue: GitHub.