SQL 2005 Cumulative Fixes

[JMast]

All,

We just upgraded from SQL 2000 to SQL 2005. We installed SP3, but none of the cumulative packages. I admit to being overwhelmed and under pressure so I didn't look in detail at the fixes. I saw a warning about the filesize growth parameters being fixed in a hotfix on SP2 and made the assumption that it would be in SP3. It is not. Hotfixes for SP2 continued after SP3 was built.

So, my question for the gurus is this: Any concerns/experiencesrecommendations with the SQL 2005 SP3 hotfixes? We are running a basic JDE implementation on 2003 Enterprise with 12 GB RAM.

I am considering either cum 7 or 8 (8 just came out in February, so that may be a little new).

Any wisdom you have to share is much appreciated.

Jer

 

[klwong]

I don't remember the exact number, but the last SQL 2005 upgrade I did, I applied SP3 with cumulative update 4 or 5. I haven't heard of any problems from them re: SQL Server since at least October.

 

[brother_of_karamazov]

[ QUOTE ]
All,

We just upgraded from SQL 2000 to SQL 2005. We installed SP3, but none of the cumulative packages. I admit to being overwhelmed and under pressure so I didn't look in detail at the fixes. I saw a warning about the filesize growth parameters being fixed in a hotfix on SP2 and made the assumption that it would be in SP3. It is not. Hotfixes for SP2 continued after SP3 was built.

So, my question for the gurus is this: Any concerns/experiencesrecommendations with the SQL 2005 SP3 hotfixes? We are running a basic JDE implementation on 2003 Enterprise with 12 GB RAM.

I am considering either cum 7 or 8 (8 just came out in February, so that may be a little new).

Any wisdom you have to share is much appreciated.

Jer

[/ QUOTE ]


I do not apply SQL hotfixes unless I am experiencing a problem that the hotfix addresses. Hotfixes from Microsoft do not receive the same level of testing as Service Packs and thus are less reliable.

Microsoft says "...this hotfix is intended to correct only the problem that is described in this article. Apply this hotfix only to systems that are experiencing this specific problem."

 

[JMast]

Ken,

That is good to hear. Thanks for the info.

Jer

 

[JMast]

Jeff,

You are getting to the reason for the post. I don't like to apply hotfixes "just because" either. I am new to Microsoft patches and the warnings about hot fixes are frequent on their site.

We technically don't have a problem right now, but if we forget and change a filesize parameter, it could hit us hard and fast so I would like it fixed. The fix first shows up in Cum 1, so an argument can be made to stop there.

However, we have another annoying issue in JDE that showed up after the upgrade that is very slippery to track, so I was thinking of using the shotgun approach hoping a later update makes it disappear.

The issue is with P4210 blowing up Sales Orders. It is very random and I have had no luck tracking it down yet.

Jer

 

[klwong]

Amazingly enough, the JDE MTR page for SQL 2005 and SQL 2008 states:

[ QUOTE ]
Applying the Latest Available Cumulative Update From Microsoft is Recommended

[/ QUOTE ]

 

[JMast]

Nice catch, Detective Ken.

I did not get quite that stance from Oracle today:

"We do certify Service Packs on SQL Server. However, we do not validate Cumulative Update Packages. Our position is that a customer can apply update packages if they feel there is benefit."

Well, I guess this is what backup tapes and a coin are for... Heads update 1, Tails update 8....

Jer

 

[brother_of_karamazov]

[ QUOTE ]
Amazingly enough, the JDE MTR page for SQL 2005 and SQL 2008 states:

[ QUOTE ]
Applying the Latest Available Cumulative Update From Microsoft is Recommended

[/ QUOTE ]

[/ QUOTE ]

Of course they recommend that, it's not their software. I am sure that Microsoft recommends you are completely up-to-date on application code too, regardless of the instability implications. It's an easy thing to to do recommend someone else's code be applied - your butt is covered and any adverse impact can be blamed on the other buy's code.

Heck, here's a recommendation from Jeff Stevenson:

Apply every single code fix from every vendor as soon as it comes out. However, do not call me if something breaks, I'm just trying to make sure everyone is up-to-date.

 

[Terry DeMoure]

Hey Jer

The P4210 problem is the one I have been working on getting resolved. Please mention SAR 8955985 in your SR you open with Oracle Support on the problem with the P4210 blowing up which is what is happening to us and dozens of other Oracle customers.

[ QUOTE ]
The issue is with P4210 blowing up Sales Orders. It is very random and I have had no luck tracking it down yet.
Jer

[/ QUOTE ]

 

[JMast]

Terry,

Thanks for the heads up. Should I be able to see the SAR on Oracle? When I search on that number all I get is a bug related to IE8. I haven't searched for a specific SAR in years, so I may be doing something wrong.

Thanks,

jer

 

[Terry DeMoure]

Actually you did find the right thing. The P4210 blowing up problem is a direct result of the bug involving E1, TR 8.98.2, and IE8. On larger sales orders, drilling down caused the application to stop and begin flashing or close uncontrollably.

I had to re-learn how to search SARs today because the Oracle Support Tech (when he closed our SR as Problem Resolved...was that a joke? lol) he told me to login to My Oracle Support and search daily for the progress on the SAR by seaching in Patches and Fixes, which found zero results. It wasn't until this morning that I learned from my JDEList contemporaries to search for the SAR (by number only) in the Knowledge Browser.

Here is the link:

SAR 8955985

 

[JMast]

Terry,

That SAR is not related to our issue. We are in the stone age of ERP 8.0 fat clients. The error occurs when the CSR clicks Place Order.

Jer

 

[Terry DeMoure]

Oops...I have tio get used to looking at the bottom of everyone's signature to see their instances and versions. I guess that I only focused on "P4210 blowing up" and thought that related to our "explosive" problem too.

 

[jacksmccu]

Hi Jer,

Cumulative updates for SQL Server are released in general every 2 months and contain a limited number of hotfixes that have been made available for individual download prior to its inclusion in the CU. If a hotfix is included in a CU that means it will also be included in the next Service Pack. There are separate CU streams, one for each supported service pack, so as users are not forced to apply a service pack to fix a specific problem. The inclusion of a particulare hot fix in a CU for SP2 does not mean that it will be included in a CU for SP3 as it may not be required in SP3. I have pasted a link below that has more detail on the CU process.

http://support.microsoft.com/kb/935897

Brother of Karamazov is correct when he says CUs are not as well tested as Service Packs but in general CUs are better tested than individual hotfixes. This is true for all software vendors, patches or PTFs are not as well tested as full or incremental releases. That does not mean you should apply them blindly, hence the warning on the KB article. However if you either have a specific issue mentioned in the CU or you think could be prone to an included issue it is certainly worth testing. Neither Microsoft or Oracle recommend applying patches directly to a running Production instance. All patches, from whichever vendor, should be tested in your environment prior to being applied in Production.

All that said I think you have a specific concern over a hotfix that is in a CU for SP2 but not in SP3. This usually means that the hotfix is not required on the later SP but if you let me know the hotfix or KB number for the fix you are concerned about I will look at.

Hope this helps,

Jack.

Ps. For those who haven't guessed already or aren't sure yes I work for Microsoft and have been working on JDE for over 12 years now.