Error - “Unexpected. Please try again.” in BigQuery

Question

I am trying to move data from one table to another by using "select *" and "--destination_table" option, but I keep getting the "Unexpected" error. Trying the same query with a where clause, which limits the data to a five second interval, results in success.

Failed JobId - divine-builder-586:bqjob_r7fe3d3a191881823_0000014662d302c5_1

Successful JobId - divine-builder-586:bqjob_r7f879846473f9b08_00000146631d5857_1

Any Ideas?

The query which succeeds is as follows :

Select tstamp,uniq_id,INTEGER(ip) as ip,devtype, FLOAT(price) as price,FLOAT(altitude) as altitude,FLOAT(longitude) as longitude,FLOAT(latitude) as latitude FROM [bigquery_processed.table_rc2] WHERE tstamp >='2014-05-01 00:00:00' and tstamp <'2014-05-01 00:00:05'

Job ID: divine-builder-586:bqjob_r7f879846473f9b08_00000146631d5857_1

Start Time: 12:33am, 4 Jun 2014

End Time: 12:41am, 4 Jun 2014

Bytes Processed: 22.8 TB

Destination Table: divine-builder-586:bigquery_processed.table_rc3

Write Preference: Append to table

Allow Large Results: true

The query which fails does not have where clause, rest is same.

Answer 1

This is what I got to know - Unexpected error usually occurs when there is some internal issue in BigQuery. Most of the times running the job again after some duration fixes the issue. But some times Google needs to debug and fix the issue(if you contact their support team)

Hopefully it helps someone :)

Answer 2

One workaround is to split the data into smaller parts using a WHERE clause with the mod function.

Try this?

Select tstamp,uniq_id,INTEGER(ip) as ip,devtype, FLOAT(price) as price,FLOAT(altitude) as altitude,FLOAT(longitude) as longitude,FLOAT(latitude) as latitude FROM [bigquery_processed.table_rc2] WHERE ABS(HASH(devtype) % 2) = 0

And then append the table with same query, just with WHERE ABS(HASH(devtype) % 2) = 1.

Honestly, this should only be a temp workaround. It should be able to run your original query since there are no JOINs or GROUP BYs with huge buckets.