Informix

Two questions:

1. I've used sqltrace for many years but have one area that I can't get clear ... sql_uid mapping to users via sql (not SS, OAT or HQ). I capture via a scheduler task the content of sysmaster:syssqltrace to sysadmin:mon_syssqltrace. sql_uid equates to a number and what I'm looking for is where these numbers are mapped. I know they correlate with syssessions, but what about users that don't have a session in syssessions? 

2. I've turned on sqltrace at the "user" level and then added a user, let's say "skippy" to the user list. But I still am getting other users as well (agreed - this may be leftovers but my trace buffer duration is 5 secs so I doubt it). I may be getting the order of ops wrong since I've typically used SS or OAT for this over the years. 

Mark:

You are overthinking this one. The sql_uid is the OS user id for the connected user:

informix@Elezar-II:~$ dbaccess sysmaster -
> select * from syssqltrace where sql_database = 'art';

sql_id            65
sql_address       1501114952
sql_sid           90
sql_uid           1000
sql_stmttype      1
sql_stmtname      DATABASE
sql_finishtime    1679412473
sql_begintxtime   1210465378
sql_runtime       0.00246567244
...

art@Elezar-II:~$ id
uid=1000(art) gid=1000(art) groups=1000(art),4(adm),24(cdrom),27(sudo),30(dip),46(plugdev),122(lpadmin),133(lxd),134(sambashare),1001(informix)

Two questions:

1. I've used sqltrace for many years but have one area that I can't get clear ... sql_uid mapping to users via sql (not SS, OAT or HQ). I capture via a scheduler task the content of sysmaster:syssqltrace to sysadmin:mon_syssqltrace. sql_uid equates to a number and what I'm looking for is where these numbers are mapped. I know they correlate with syssessions, but what about users that don't have a session in syssessions? 

2. I've turned on sqltrace at the "user" level and then added a user, let's say "skippy" to the user list. But I still am getting other users as well (agreed - this may be leftovers but my trace buffer duration is 5 secs so I doubt it). I may be getting the order of ops wrong since I've typically used SS or OAT for this over the years. 

Thanks Art. I haven't "overthinked" (sic) anything since uh, yesterday. 

So for windows client or a web app coming in with a connection string/passwd embedded - where does UID come from in those cases? 

Any thoughts on Q#2 n original thread? May just be a syntax issue ... if so, nevermind, I'll work through it eventually. 

Thanks! 

Mark:

You are overthinking this one. The sql_uid is the OS user id for the connected user:

informix@Elezar-II:~$ dbaccess sysmaster -
> select * from syssqltrace where sql_database = 'art';

sql_id            65
sql_address       1501114952
sql_sid           90
sql_uid           1000
sql_stmttype      1
sql_stmtname      DATABASE
sql_finishtime    1679412473
sql_begintxtime   1210465378
sql_runtime       0.00246567244
...

art@Elezar-II:~$ id
uid=1000(art) gid=1000(art) groups=1000(art),4(adm),24(cdrom),27(sudo),30(dip),46(plugdev),122(lpadmin),133(lxd),134(sambashare),1001(informix)

Two questions:

1. I've used sqltrace for many years but have one area that I can't get clear ... sql_uid mapping to users via sql (not SS, OAT or HQ). I capture via a scheduler task the content of sysmaster:syssqltrace to sysadmin:mon_syssqltrace. sql_uid equates to a number and what I'm looking for is where these numbers are mapped. I know they correlate with syssessions, but what about users that don't have a session in syssessions? 

2. I've turned on sqltrace at the "user" level and then added a user, let's say "skippy" to the user list. But I still am getting other users as well (agreed - this may be leftovers but my trace buffer duration is 5 secs so I doubt it). I may be getting the order of ops wrong since I've typically used SS or OAT for this over the years.