Skip to Content
avatar image
Fernando Sánchez Sánchez

Could anyone write an example of then -f option of hdbsql?

Jun 18, 2018 at 07:56 AM | 159 Views | Last edit Jun 18, 2018 at 08:44 AM 2 rev.

Hi

I need to execute batch process in hdbsql and I need a log file to check what was wrong

I've read about -f option: returns all SQL statements that are sent to the database instance

That's exactly what I want, but I don't know how to use this option.

Could you please help me?

SAP HANA
hdbsql | options
Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

1 Answer

  • avatar image
    Lars Breddemann
    Jun 18, 2018 at 12:00 PM

    Not quite sure that this feature will help you a lot, but hey, you asked for examples, so take these:

    No option:

    hdbsql -U s12dev "SELECT count(*) from fact"
COUNT(*)
10000000
1 row selected (overall time 574.115 msec; server time 268 usec)

    -f option (printout the SQL commands):

    hdbsql -U s12dev -f "SELECT count(*) from fact"
SELECT count(*) from fact

COUNT(*)
10000000
1 row selected (overall time 1301.303 msec; server time 139.424 msec)

    -fn option (printout the SQL commands with additional line numbering):

    hdbsql -U s12dev -fn "SELECT count(*) from fact"
1) SELECT count(*) from fact

COUNT(*)
10000000

1 row selected (overall time 606.342 msec; server time 178 usec)

    -x option (run quietly (no messages, only query output)):

    hdbsql -U s12dev -fn -x "SELECT count(*) from fact"
COUNT(*)
10000000

    Add comment
    10|10000 characters needed characters exceeded

    • Lars Breddemann Fernando Sánchez Sánchez
      Jun 18, 2018 at 03:06 PM

      Alright, I don't care how you build up the script that you eventually call.

      Show the final script and I look into why it fails.

      Some general points, though:

      • passing in username/password as clear text (it will be in the final script) is a serious security bug. Do, what I've shown above and use hdbuserstore instead.
      • there is no need to redirect stdout. hdbsql provides the -O parameter to define an output file.
      • not sure why you want to specify the multi-line mode while in interactive mode. simply specifying the command line option -m would suffice.
      • same goes for the input file: use the -I parameter instead
      • the -fn options is only available as of HANA 2 SP 03. If your hdbsql client is on an older version (try hdbsql -v to display the version), then these options don't exist.

      Anyhow, even on Linux the following produces an output file that includes the SQL statements:

      hdbsql -U ddude -f -o /tmp/out.txt "SELECT count(*) from fact"

cat /tmp/out.txt
SELECT count(*) from fact

COUNT(*)
10000000

In case of an error, this is also recorded:

hdbsql -U ddude  -f -o /tmp/out.txt "SELECT count(*) from factX"
* 259: invalid table name:  Could not find table/view FACTX in schema DEVDUDE: line 1 col 22 (at pos 21) SQLSTATE: HY000

cat /tmp/out.txt
SELECT count(*) from factX

* 259: invalid table name:  Could not find table/view FACTX in schema DEVDUDE: line 1 col 22 (at pos 21) SQLSTATE: HY000