Home/ Questions/Q 8445919
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-10T09:47:10+00:00

I am running Sphinx 0.9.9-release (r2117). I get the following error, when I run

  • 0

I am running Sphinx 0.9.9-release (r2117).

I get the following error, when I run the indexer, which reads data from a MySQL table:

indexing index 'main'...
ERROR: index 'main': ., 888.4 MB
total 3629180 docs, 888448551 bytes
total 170.437 sec, 5212749 bytes/sec, 21293.30 docs/sec

As you can see, the indexer simply stops at 888.4 MB without a real error message.
The same happens even if I delete the complete index and try to create a new one. If I do that, the newly created index files are empty.

The system worked for months now, but the row count of the table has increased from about 11 million to about 24 million. Also it must have really huge id gaps, because minimal id is 1 and maximum id is > 6 billion.
I suspected there might be something wrong with MySQL, so I created a small script which executes all “paging” queries, that should be executed by Sphinx during indexing:

SELECT ... FROM ... WHERE id >= $start AND id <= $end

The script executed all queries for alle id ranges without returning an error.

Some maybe relevant Sphinx configuration details:

sql_query_range         = SELECT MIN(id), MAX(id) FROM offer
sql_range_step          = 10240

SELECT o.id, ....
            FROM offer o \
            LEFT JOIN customer c ON c.id = o.customer_id \
            WHERE ... AND o.id >= $start AND o.id <= $end AND o.last_update <= (SELECT last_full_index FROM index_update WHERE name = 'default')

Is there some way to increase verbosity of the indexer?

Any ideas how I could narrow the problem?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    0.9.9 is REALLY old now. Lots of bug fixes and improvements, since then.

    Just upgrading may well fix your issue.

    But it also brings you
    –print-queries
    which may be useful.

    Or even –dump-rows – but make sure you have a big temporally location to put it. Tailing it when get the error might help.

    Another thing to check is wait_timeout on your mysql server. Maybe there is just too long between queries, that mysql is killing the connection

    • 0