[R] unixodbc - postgresql: broken pipes, malloc, ksqo and other errors

Duane Winner duanewinner at att.net
Fri Mar 5 15:34:02 CET 2004


Hello all. I am posting this on unixodbc-dev and pgsql-odbc as well as
r-help to make sure all of my bases are covered (and because I'm not
sure which one of these is the source of the issue.)

Unixodbc with Postgresql seems to be working, but I am getting some
errors that I don't understand and want to know why I am getting these
errors before I put everything into production.

I am running:
FreeBSD 4.9-RELEASE
unixODBC-2.2.8 (installed from FreeBSD port)
postgresql-7.3.5_1 (installed from FreeBSD port)
R 1.8.1 (installed from R-letter FreeBSD port)
RODBC (installed from CRAN)

Here is what is happening:

When I run "isql":
# isql mydb -v
+---------------------------------------+
| Connected!                            |
|                                       |
| sql-statement                         |
| help [tablename]                      |
| quit                                  |
|                                       |
+---------------------------------------+
SQL> q
Broken pipe

*** Why am I getting "broken pipe"? What does it mean and does it
matter?


Also in R:
> library(RODBC)
> channel <- odbcConnect("mydb")
> weather <-sqlQuery(channel,"select * from weather")
> odbcClose(channel)
R.bin in free(): warning: chunk is already free

*** Why am I getting this warning? This one concerns me the most, as I
need to make sure that it doesn't cause problems on a production system.
I don't understand much about programming, but I have been told that
this is essentially a 'malloc' error. 


Also, in ANY application (isql, OpenOffice, R) that opens a connection
using unixODBC, I see in my postgresql log this message:

ERROR:  'ksqo' is not a valid option name

*** Why am I getting this error, and again, should I be concerned?
Google actually gave me some results on this one, and to the best of my
discerning, it looks as if it is some Microsoft query option that just
isn't supported, in which case, fine. I'm just looking for assurances.


Again, like I said, I am able to run queries and utilize unixodbc and
the postgresql driver as best as I can tell. But I would hate to
implement this into production and find out that there is some critical
problem that will bite me once the production apps start to take some
hits.

Thanks for any info/assistance/guidance you can provide.

-DW




More information about the R-help mailing list