• Article Type: General
• Product: Aleph
• Product Version: 20

Description:
Looking in util o/14/1 we see the message: "TABLESPACE_WITHOUT_FREE_SPACE" for a particular tablespace. How should we handle this?

Resolution:
If this tablespace is "autoextend", then Oracle should automatically create a datafile.

Otherwise, you or your DBA should add a datafile to this tablespace using util o/13/3 or another method ASAP.

What you do next depends on whether the tablespace name is "TSnD" or "TSnX" (where "TSn" is TS1, TS2, TS3, or TS4).

A TSnX tablespace has Oracle indexes only in it. The following SQL shows what Oracle indexes are in a tablespace:

SQL> select index_name from all_indexes where tablespace_name = 'TSnX' order by index_name;

If you have seen errors in the log or odd behavior in connection with a particular table, then, after space is added to the tablespace, you should do util a/17/3 to rebuild that table's indexes.

At some point (the following night?) you should do util a/17/3 for all of the indexes found by the above SQL. You could, instead, do util a/17/3 specifying "All". That would take more resources, but will also work.

A TSnD tablespace has Oracle tables (data) in it. If there have been errors in writing data records, that *should* have been apparent from logs or from messages which users received.
If the table is an Aleph index table, then you may want/need to run the index job to recreate it. (See KB in regard to what tables are index tables.) KB 8192-2133 describes how to locate and re-index bib records updated since x date; that can make a run of the complete indexing job(s) unnecessary.
If the table is a "corpus" table, then there is no way to recover the updates which failed due to the lack of space.

• Article last edited: 10/8/2013