Oracle Discovery

Storage error during script-based discovery

This error has occurred during ASD but the assumption is that it could also occur during script-based discovery and even potentially during SSH-based legacy discovery. TADDM runs the query "select file_name, tablespace_name, to_char(bytes) from dba_data_files" during discovery. Sometimes the output of this can be incorrect, where the file_name contains a special character and causes an extra line. See the following sample.

/u02/database/db\_hdf/hdfi/datafile/infra/audit\_archive\_t\_i1005\_1005\_ts01.ora
<#>AUDIT\_ARCHIVE\_I1005\_1005      <#>10485760  
/u02/database/db\_hdf/hdfi/datafile/db\_admin/index/audit\_archive\_i1006\_1006\_1.or
<#>AUDIT\_ARCHIVE\_I1006\_1006      <#>10485760  
a
<#>                                 <#>

Notice that the 'a' was cut off the end of the previous file_name and has it's own line with a blank tablespace_name and bytes. This will cause TADDM to have a storage error during discovery. The only way to fix it is to alert the DBA of the issue and have him or her correct it on the database side. As a workaround during ASD, you can manually edit the output and delete that line.

Sep 2015 Update: APAR IV77128 resolves this issue. This fix won't make it into v7.3.0.2 but it is safe to assume it will make it into v7.3.0.3.