nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

iw_cgi_wrapper.cgi error, even after sticky bit setting

FanTom

Hello,

I am stuck with the workflow starting this weekend, as the

iw_cgi_wrapper.cgi was

-rwsrwxr-t 1 root daemon 3690180 Feb 11 2007 iw_cgi_wrapper.cgi

So I changed (using sudo su ) to

-rwsrwxr-x 1 root other 3690180 Feb 11 2007 iw_cgi_wrapper.cgi

I still have the same error. Any pointers?
I am getting Internal server error soon after I start the Job, from my WF debug window.

Also to note that if there are other cgi's in the same folder as this wrapper - do they all need to be set with same permissions as this? I mean, the custom written cgi's?

Thanks,
N/-.,
----------------------------------------
Solaris 2.9, TS 6.5 OD 6.0

Find more posts tagged with

Comments

nipper

did someone by chance remove the folder iwhome/tmp/cci ?

Do any WFTs work ?

Has the WFT changed recently ?

FanTom

Hello Nipper,
Thanks for reply.

Yes, I do see the cci folder newly created today.
Even after it is created, same error after I start the job.

drwxrwxrwx 2 root root 8192 Mar 1 13:57 cci

So, I suppose it was removed and is newly created. Its blank folder.

Please advise.
N/-.,

nipper

All WFTs ? Or just one failing ?

region.rptdesign

FanTom

Further investigation, shows its only one WFT for particular category.

Another cgi is making a connect to DB.

What it seems like is when this CGI runs and impersonates the logged in user to connect to DB, it cannot continue.

Even TNS ping as that logged in user fails connect.

Now, if I do not want the wrapper CGI to not impersonate and connect to DB with root credentials, could I set iw_cgi_wrapper.cgi to be owned and executed by root? Which means not setting the uid.

Or any other better ways.

N/-.,
----------------------------------------
Solaris 2.9, TS 6.5 OD 6.0

Adam Stoller

Further investigation, shows its only one WFT for particular category.

Another cgi is making a connect to DB.

What it seems like is when this CGI runs and impersonates the logged in user to connect to DB, it cannot continue.

Even TNS ping as that logged in user fails connect.

Now, if I do not want the wrapper CGI to not impersonate and connect to DB with root credentials, could I set iw_cgi_wrapper.cgi to be owned and executed by root? Which means not setting the uid.

Or any other better ways.

N/-.,
----------------------------------------
Solaris 2.9, TS 6.5 OD 6.0

If you do NOT want impersonation - do NOT use iw_cgi_wrapper.cgi.

If your CGI script is throwing an internal error - look through the IWHOME/local/logs/iwui/servletd_out.log and/or IWHOME/local/logs/iwutild_cmdout.log for potentially more detailed information - it could be that your program is printing something to STDOUT before the initialization of the stream ("content-type: text/html\n\n")

Assuming this is a Perl script - you can also try adding:

use CGI::Carp qw(fatalsToBrowser);

which may help you see more details of the error in the browser window itself.

XTabColumnStyling.rptdesign

FanTom

Ghoti,
Thanks for this info. I will use this and verify.

However -- as in the corporate environment - being informed is the Key.
It turns out that the db instance I should be talking to is down for routine maintenance and so the custom cgi is not able to read values from DB.

An old script with no appropriate message logging, added with disconnected notification of different pieces.(^$(&$ wonder puts us into situations cracking heads.

Thanks guys for the pointers and help.
Hope as soon as db is up, I wont see any problem. X X

N/-.,
---------------------------------------------------------------
Solaris 2.9, TS 6.5, OD 6.0. Oracle 8i with ver 9 Client on TS box.

Adam Stoller

I'd suggest fixing the script to check for errors and provide useful messaging - doing it while the DB is down is the best time to verify ;-)