The following bug has been logged on the website:
Bug reference: 11989
Logged by: Antonio Galery
Email address: galery@transfacil.com.br
PostgreSQL version: 9.2.9
Operating system: AIX 7100-03-03-1415
Description:
At 14 November at 23:38 postgresql log records a checkpoint starting and
does not register the checkpoint completed. No other checkpoint was
conducted after the beginning of the checkpoint 23:38. At 23:58 the database
deactivated with the following message: 14/11/2014 23:56:39 GMT + 3
[17,825,796] - [10.57.1.26] [585-1] citgis = db, user = opegis STATEMENT:
SELECT * FROM pcd_22_processamento ($ 1)
14.11.2014 23:58:46 GMT + 3 [12,320,794] - [10.57.1.26] [1-1] = citgis db,
user = opegis PANIC: could not write to file "pg_xlog / xlogtemp.12320794":
No space left on device. Report that the area had pg_xlog size of 7GB but
occupied until then 4gb. Doubled the area in order to reactivate the
database. I found no explanation for this.
Galery.
