DST 2007 - CF7

[AdobeCF7User]

I've looked at the various kb articles, but I'm not sure what I need to do.

It appears Java needs to be bumped up to 1.4.2_11 for DST 2007, but the 7.0.1
and 7.0.2 updaters don't appear to take care of that.

Adobe recommends you only use a 7.0.2 updater if you have a specific issue
resolved by it and the release notes say it will only get you to Java _9.
However, this tech note ([url]http://www.adobe.com/go/d2ab4470[/url]) indicates 7.0.2 has
been certified for 1.4.2_11. Does it only install _9 but has been tested to
work with _11?

The current version I have installed is:

Server Product / Version: Cold Fusion MX 7,0,1,116466
Java Version: 1.4.2_05
Installation folder: custom path

So is there a definitive course of action to update CF7 for DST 2007?

Are the old versions of Cold Fusion (4.0, 4.5, 5.0) affected by DST 2007?

[tmschmitt]

Have you looked at:
([url]http://www.adobe.com/go/d2ab4470[/url])
Upgrading to 1.4.2_11 was the path that I took for CF7.

I know the CF5 has issues with the new DST as well. Specifically the
DateConvert() function as near as I can tell.
[url]http://www.adobe.com/go/kb400365[/url] covers CF5, but I have my doubts on the
sample code provided and have commented on Sarges Blog about it. I'm still
working on a solution for that.

[AdobeCF7User]

Hey Terry, I had posted at the same time as you on Sarge's blog asking about 4.5! So you just updated to 1.4.2_11 and nothing else and everything is running ok?

[tmschmitt]

1.4.2_11 was all I installed and it seems to be working fine. Now keep in mind
that I am not using CF7 much. I unfortunately still use CF5 from most apps.
Install it on a test machine first and try it out. The original JRE will still
be under the CF folders, so you can always revert back if you do have trouble.
I'm still hoping for some feedback at Sarge's blog on my CF5 testing results.
The code in the KB for CF5 also contains syntax errors and incorrect variable
names, so it wouldn't even execute, but I'm more concerned about the results of
the working code.

Terry

[AdobeCF7User]

Terry, once patched CF5 should be the same as CF7 with the exception of the dateconvert function so did you use the same code to test both servers? Can you post the code you used to test? Thanks

[tmschmitt]

Here is the test page. I describe the test over at Sarge's blog. The
conversion code is copied directly from the blog with only the input date
object changing. Instead of using now(), I am passing in a date.

Terry

----------------------------------------------------------

<h1>Date UDF Test (Patched OS using Sarge's code)</h1>
<h2>Server Time: <cfoutput>#now()#</cfoutput></h2>

<!---
Modify this date for testing. The input date is an assumed UTC date.
--->
<cfset utcTime = CreateDateTime(2007, 2, 15, 8, 0, 0) />
<cfset cfTZ = getTimeZoneInfo() />
<cfset dateObj = utcTime />
<cfif cfTZ.isDSTOn>
<cfset utcDate = dateAdd('h', -1, dateConvert("Local2UTC", dateObj)) />
<cfset localDate = dateAdd('h', 1, dateConvert("UTC2Local", dateObj)) />
<cfelse>
<cfset utcdate= dateConvert("Local2UTC", dateobj) />
<cfset localDate = dateConvert("UTC2Local", dateObj) />
</cfif>

<hr>
<cfdump var="#cfTZ#">
<h3>UTC time (Input): <cfoutput>#utcTime#</cfoutput></h3>
<h3>Local time (Output): <cfoutput>#localDate#</cfoutput></h3>

<!--- Results:
Scenario 3 - Patched OS using Sarge's code
Server: 2/15/2007
UTC: 2/15 8hr diff normal
UTC: 3/15 8hr diff did not adjust**************
UTC: 4/15 7hr diff adjusted
Server: 3/15/2007
UTC: 2/15 7hr diff invalid adjust**************
UTC: 3/15 7hr diff adjusted
UTC: 4/15 6hr diff invalid adjust**************
Server: 4/15/2007
UTC: 2/15 7hr diff invalid adjust**************
UTC: 3/15 7hr diff adjusted
UTC: 4/15 6hr diff invalid adjust**************
--->

[AdobeCF7User]

Thanks Terry, I'm going to patch this weekend and give it a go!

[Brad Melendy]

I just did my patch, and thankfully I did it on a test instance because it
killed CF entirely. I edited back the jvm.config file and it still won't
work? How can the reliance on JAVA be so fragile?? I'm wondering if I'll
be able to apply the patch at all now because it can't go on production
servers if this is the result. Sigh..

"AdobeCF7User" <webforumsuser@macromedia.com> wrote in message
news:eqh3sf$f8o$1@forums.macromedia.com...

> Thanks Terry, I'm going to patch this weekend and give it a go!

[PaulH **AdobeCommunityExpert**]

Brad Melendy wrote:

> I just did my patch, and thankfully I did it on a test instance because it
> killed CF entirely. I edited back the jvm.config file and it still won't
> work? How can the reliance on JAVA be so fragile?? I'm wondering if I'll
> be able to apply the patch at all now because it can't go on production
> servers if this is the result. Sigh..

did you d/l & install the server version of the JDK? are you pointing at right
place? is this standalone cf or multi-server?

[Brad]

Did all that, I'm using Fedora Core 6 and CF 7.02 to test this patch
process.

To select the patch, I sent to the link in Adobe's tech note located here:

[url]http://www.adobe.com/cfusion/knowledgebase/index.cfm?id=d2ab4470[/url]

and selected J2SDK/J2RE - 1.4, from there I had two options, 1. SDK and 2.
JRE. I choose the SDK which was the larger of the two. Sun doesn't seem to
list anything they call the "Server" Version of anything.

After switching to the new Java version and pointing the jvm.config file to
the new location, CF says it has started in the logs, but with one error:

"Unable to initialize FlexAssembler service:..."

And I can't get into the CF Administrator anymore. Editing the jvm.config
file back to the old location for the JRE located inside the CF Directory
structure doesn't fix anything?? Very strange since to CF's perspective,
that jvm.config file should have been the only change.

So I'm pretty sure I've done this all correctly. Also, under windows (we
have a couple of windows servers with CF as well, but most are Linux) the
patch applied just fine, no hassles. It only seems to be a problem with
Linux. :-(



"PaulH **AdobeCommunityExpert**" <paul@sustainableGIS.com> wrote in message
news:esle7h$fd7$2@forums.macromedia.com...
Brad Melendy wrote:

> I just did my patch, and thankfully I did it on a test instance because it
> killed CF entirely. I edited back the jvm.config file and it still won't
> work? How can the reliance on JAVA be so fragile?? I'm wondering if I'll
> be able to apply the patch at all now because it can't go on production
> servers if this is the result. Sigh..

did you d/l & install the server version of the JDK? are you pointing at
right
place? is this standalone cf or multi-server?

[PaulH **AdobeCommunityExpert**]

Brad wrote:

> and selected J2SDK/J2RE - 1.4, from there I had two options, 1. SDK and 2.
> JRE. I choose the SDK which was the larger of the two. Sun doesn't seem to
> list anything they call the "Server" Version of anything.

sorry, i meant SDK which has the "-server" switch option.

> structure doesn't fix anything?? Very strange since to CF's perspective,
> that jvm.config file should have been the only change.

did you stop & start the cf server service after this?

> So I'm pretty sure I've done this all correctly. Also, under windows (we
> have a couple of windows servers with CF as well, but most are Linux) the
> patch applied just fine, no hassles. It only seems to be a problem with
> Linux. :-(

i don't see any notes specific to linux about this but we don't deal w/it very
much. have you looked at steve erat's talkingtree blog entry about DST (sorry
don't have the URL handy) or maybe look thru this:

[url]http://www.stillnetstudios.com/2007/02/24/upgrading-the-coldfusion-jvm-on-linux-and-windows/[/url]

[PaulH **AdobeCommunityExpert**]

i actually meant this entry:

[url]http://www.talkingtree.com/blog/index.cfm/2006/4/7/CFAdmin-problems-when-editing-JVM-settings[/url]

[Brad]

Yes, tried restarting CF numerous times and then rebooted the server as
well. I think it is a versioning issue. The difference between Linux and
Windows (that we are painfully running into now, first with MySQL and now
with CF) is that Linux is versioned so often that often times things just
won't work unless you have the correct conbination of elements in Linux
first. Eveyrthing just seems to work so much better under Windows and maybe
that is the answer.

"PaulH **AdobeCommunityExpert**" <paul@sustainableGIS.com> wrote in message
news:esn1lp$6mu$1@forums.macromedia.com...
Brad wrote:

> and selected J2SDK/J2RE - 1.4, from there I had two options, 1. SDK and 2.
> JRE. I choose the SDK which was the larger of the two. Sun doesn't seem
> to
> list anything they call the "Server" Version of anything.

sorry, i meant SDK which has the "-server" switch option.

> structure doesn't fix anything?? Very strange since to CF's perspective,
> that jvm.config file should have been the only change.

did you stop & start the cf server service after this?

> So I'm pretty sure I've done this all correctly. Also, under windows (we
> have a couple of windows servers with CF as well, but most are Linux) the
> patch applied just fine, no hassles. It only seems to be a problem with
> Linux. :-(

i don't see any notes specific to linux about this but we don't deal w/it
very
much. have you looked at steve erat's talkingtree blog entry about DST
(sorry
don't have the URL handy) or maybe look thru this:

[url]http://www.stillnetstudios.com/2007/02/24/upgrading-the-coldfusion-jvm-on-linux-and-windows/[/url]

[Brad]

I've got good news, when I apply the patch to an older version of linux, RH
ES3 (I know that is like 3 years old) everything works as it should.
Anything newer failed for me and I think it is the 2.6 kernel that is at the
root of things. Anyway, our production environment (being very old) is on
ES3. Our TEST/DEV servers are a little newer but it isn't critical they
have the update at this time. Thanks for everyone's help.

....Brad

"PaulH **AdobeCommunityExpert**" <paul@sustainableGIS.com> wrote in message
news:esn1od$6mu$2@forums.macromedia.com...
i actually meant this entry:

[url]http://www.talkingtree.com/blog/index.cfm/2006/4/7/CFAdmin-problems-when-editing-JVM-settings[/url]