prestodb/presto · error · PrestoException

HIVE_CANNOT_OPEN_SPLIT

HIVE_CANNOT_OPEN_SPLIT

Error message

Error opening Hive split %s (offset=%s, length=%s) using %s: %s

What it means

A generic failure while opening a Hive split's input format: any IOException from createRecordReader that is not the line-length case is wrapped as HIVE_CANNOT_OPEN_SPLIT, including the split path, offset, length, input format class, and underlying message. It means the connector could not instantiate or initialize the record reader for that split.

Source

Thrown at presto-hive/src/main/java/com/facebook/presto/hive/HiveUtil.java:323

            int headerCount = getHeaderCount(schema);
            //  Only skip header rows when the split is at the beginning of the file
            if (start == 0 && headerCount > 0) {
                Utilities.skipHeader(recordReader, headerCount, recordReader.createKey(), recordReader.createValue());
            }

            int footerCount = getFooterCount(schema);
            if (footerCount > 0) {
                recordReader = new FooterAwareRecordReader<>(recordReader, footerCount, jobConf);
            }

            return recordReader;
        }
        catch (IOException e) {
            if (e instanceof TextLineLengthLimitExceededException) {
                throw new PrestoException(HIVE_BAD_DATA, "Line too long in text file: " + path, e);
            }

            throw new PrestoException(HIVE_CANNOT_OPEN_SPLIT, format("Error opening Hive split %s (offset=%s, length=%s) using %s: %s",
                    path,
                    start,
                    length,
                    getInputFormatName(schema),
                    firstNonNull(e.getMessage(), e.getClass().getName())),
                    e);
        }
    }

    public static void setReadColumns(Configuration configuration, List<Integer> readHiveColumnIndexes)
    {
        configuration.set(READ_COLUMN_IDS_CONF_STR, Joiner.on(',').join(readHiveColumnIndexes));
        configuration.setBoolean(READ_ALL_COLUMNS, false);
    }

    public static Optional<CompressionCodec> getCompressionCodec(TextInputFormat inputFormat, Path file)
    {
        CompressionCodecFactory compressionCodecFactory;

View on GitHub (pinned to 55bb57d202)

Solutions

1. Verify the file exists and is readable at the reported path (hdfs dfs -ls / aws s3 ls) and check its integrity
2. Confirm the file's actual compression matches its extension/format declared by the table
3. Re-run the query — transient S3/HDFS errors may resolve; otherwise restore/reproduce the data
4. Check the wrapped cause message and the input format name in the error for the true root cause

Example fix

// before
CREATE TABLE t (x bigint) WITH (format='TEXTFILE', external_location='s3://bucket/parts') -- files named .gz but not gzip-compressed
// after
recompress files properly or drop the .gz suffix and set format/compression accordingly

if (!fileExists(path) || !isReadable(path)) throw new IllegalStateException("split file missing/unreadable: " + path);

try { readSplit(split); } catch (PrestoException e) { if (e.getErrorCode().getCode() == StandardErrorCode.HIVE_CANNOT_OPEN_SPLIT.getCode()) { log.warn("split open failed for {} offset {}", path, offset); /* retry with backoff, then fail with wrapped cause */ } else throw e; }

When it happens

Trigger: Opening a split whose file is missing/corrupt, wrong compression extension vs actual codec, input format class fails to initialize (misconfigured JobConf, permission denied, truncated file, checksum mismatch).

Common situations: Files deleted or moved by compaction while a query reads them; corrupt/truncated files on HDFS/S3; input format configuration errors; S3 permission or throttling issues; mismatched codec extensions (.gz file not gzipped).

Related errors

• HIVE_CANNOT_OPEN_SPLIT
• HIVE_FILESYSTEM_ERROR
• HIVE_FILESYSTEM_ERROR
• HIVE_FILESYSTEM_ERROR
• HIVE_WRITER_CLOSE_ERROR

AI-assisted analysis of prestodb/presto@55bb57d202 (2026-09-04). Data as JSON: /api/errors/62a3c2b3a633946b. Report an issue: GitHub.