But if it all checks out, then I'll look forward to hearing what the folks
say.
I'll add, as well, that your settings in the "JDBC settings" (of the
FusionReactor interface) might affect this. If you told it to show only
queries that exceed a certain number of milliseconds, it may be just that
you have none (a good thing). I can't recall if it would still show a count
of all queries to include those that did not exceed that limit. It may be
that it still would but only shows details for those that do.
/charlie
Just to be clear, simply because the datasource listing "says Fusion
Reactor" and does verify doesn't mean that the DSN is actually using the
FusionReactor (spelled without a space, btw) wrapper. Indeed, when you say
you "checked the JDBC settings" and "nothing is set there", what do you
mean? For each changed DSN, you should see it showing a JDBC URL that starts
with jdbc:fusionreactor:wrapper:{. If that's not there, check out the
FusionReactor JDBC Wrapper Tutorial and User Guides
(http://www.fusion-reactor.com/fr/support.cfm#doc) for more info.
Sorry if any of that is all obvious stuff. Since you go on to say "Is there
a chance that the database connection is still using the old
datasource without the wrapper?", I'm inclined to think you do have it all
right. Still, just pays for us to double check. :-)
As for the effect of datasource changes, they are always immediate in my
experience. No restart or cache refresh is required.
Speaking of the script, is that something you might want to share? I'm sure
others would appreciate a tool to help convert a DSN to be wrapped. I've
been meaning to develop one myself (using the Admin API in CF 7 and 8). Even
if you may prefer not to share it, how does yours work?
Before anyone gets too excited about using a tool, though, let Chris'
experience be a lesson: such an automated tool is very powerful, and it
needs to be checked (and double-checked) before and after use. I'd even
recommend taking a backup of the XML files that CF uses for the datasources
(C:\CFusionMX7\lib\neo-query.xml in CF7, and similar in 6, and
C:\ColdFusion8\lib\neo-datasource.xml in CF8, where neo-query.xml has a new
purpose.) If you mess up datasources, people (customers, colleagues, bosses,
admins, etc.) tend to get very upset!
/charlie
-----Original Message-----
From: fusion...@googlegroups.com [mailto:fusion...@googlegroups.com]
On Behalf Of Chris
Sent: Monday, June 30, 2008 2:41 PM
To: FusionReactor
Subject: FusionReactor Group: Re: Intermittent JDBC Results
As for your script to automate wrapping DSNs, thanks for sharing. It may
give folks ideas, but if I can be honest, I'd warn folks against just simply
using it. Study it for ideas, but be aware that it may not suit you.
First, he's presuming use of SQL Server, in setting port 1433. Unless all
your DSNs are using SS and are on that port, this would not be what you'd
want.
Note as well that he's also setting all the DSNs to have a single
username/password, passed in on the URL that calls the page.
Beyond the question of whether you also set userids and passwords on all
your DSNs, let alone setting them all the same, I'd also raise a caution
about using the query string (URL variables) to do that, from 2 aspects of
security.
Besides that querystring data remaining in your browser history (so someone
else on your machine could find it), the query string data also appears in
your web server (and FusionReactor) request logs. If there's any chance that
folks who see that maybe should not know that username/password (or the
cf_password which you also pass in to use for the Admin), it's a security
risk.
One solution is to use a a form which stores it in a session variable. Form
data isn't typically stored in web server (or FR) logs, and it's also not
saved in the browser history.
Not a show-stopper. Just saying it in case others would want to take ideas
from what you did. It's a useful starting point.
In fact, one last concern/observation is that I see you write over all the
current DSNs. I wonder if a better approach would be to use the old ones to
create new ones and leave the old ones as renamed. That way you'd always
have them to fall back on in an emergency. Of course, taking a backup of the
xml file first is nearly as good. Again, just sharing ideas for others, not
knocking what you've chosen to do for yourself.
/charlie
-----Original Message-----
From: fusion...@googlegroups.com [mailto:fusion...@googlegroups.com]
On Behalf Of Chris
Sent: Tuesday, July 01, 2008 10:00 AM
To: FusionReactor
Subject: FusionReactor Group: Re: Intermittent JDBC Results