I'm afraid you need to change a query, too:
In uptime-defs.tcl, you should change:
proc uptime_monitor_list_of_ids {db monitor_ids} {
...
...
set selection [ns_db 0or1row $db "select uu.*,to_char(time_when_first_unreachable,'YYYY-MM-DD HH24:MI:SS') as full_unreachable_time, to_char(sysdate,'YYYY-MM-DD HH24:MI:SS') as full_sysdate, round((sysdate - time_when_first_unreachable)*60*24) as n_minutes_downtime
...
for something like this:
...
...
set selection [ns_db 0or1row $db "select uu.*,to_char(time_when_first_unreachable,'YYYY-MM-DD HH24:MI:SS') as full_unreachable_time, to_char(sysdate(),'YYYY-MM-DD HH24:MI:SS') as full_sysdate, round(date_part('epoch', sysdate() - time_when_first_unreachable)/60) as n_minutes_downtime
from ...
since i think PostgreSQL doesn't have a functino round(interval).
Ah, you should also change every call of sysdate for sysdate() and create the usual openacs replacement for sysdate():
create function sysdate() returns datetime as '
begin
return ''now'';
end;' language 'plpgsql';
And, since i've started writing, i'll also post a possible redefinition for uptime_stale_p:
create function uptime_stale_p (datetime)
returns varchar
as '
declare
time_when_first_unreachable alias for $1;
begin
IF time_when_first_unreachable is null
THEN return ''f'';
ELSE
IF time_when_first_unreachable > (sysdate() - 10)
THEN return ''f'';
ELSE return ''t'';
END IF;
END IF;
return result;
end;
' language 'plpgsql';
Hope that helps!
