Hello, I am trying to run multiple TPT jobs in parallel(using selector operator to select data from a Teradata table that has BLOB and CLOB columns and data connector operator to write the data in files in deferred mode ), and getting the following error:
SQL_SELECTOR: TPT15128: Error: 28, while flushing buffered data to a file '/opt/teradata/client/13.10/tbuild/logs/MY_TPT_JOB_LOBFileList_p1'
Reason: No space left on device
FILE_WRITER: TPT19221 Total files processed: 0.
SQL_SELECTOR: disconnecting sessions
There is enough space on the disk, so disk space is most probably not an issue.
When researched on this error, I found the following
Cause:Error 28, ENOSPC, on a semop() means that the system has run out of undo structures for semaphores.
1 Use the sysdef command or something similar to find out the value of the semaphore undo structures (SEMMNU) defined on the system. 2 Increase the value of SEMMNU. 3 Reboot the system. 4 Re-launch the job.
Would like to understand this error and corrective action. Is it really a diskspace issue? Is it happening because too many TPT jobs are running in parallel?
Also, I have varchar LOBDIRECTORYPATH mentioned for the directory where the LOB files would be created, but as you see in the error, some big file is created in the default log dir (/opt/teradata/client/13.10/tbuild/logs/) that is not a log file. Does TPT normally create a file there when data connector operator is used? and How to avoid creating this file?
Would really appreciate your reply on these questions.