SP2 fallout : the drama continues; UPDATE 4/6
4/6
Here is some more info from Steve Kass. He paraphrased Microsoft's internal response to the confusion about who should install 3054 and who should install 3159.
<snip>
3054 includes the new fix along with a few important general-release fixes since SP2. But 3054 doesn't include several quick fix updates that weren't general-release, but that some customers have needed and installed. These latter customers can't install 3054, because it doesn't have the newest quick fixes they need. So they get a different build–3159–one that contains the newest fix, and also contains the post-3054 quickfixes.
Build 3159 isn't for everyone just because it contains more fixes. The extra fixes it contains were not tested (maybe not even designed) to be general-distribution fixes, and it would be risky if not a plain-out bad idea to distribute them generally, along with the current critical fix.
</snip>
——————————-
Steve Kass told us a short while ago that there are two new hotfixes available to finally correct the maintenance plan issues brought about by SP2 and all of its descendants.
Check @@VERSION.
If SQL Server is between 3042 and 3053, get build 3054 from http://support.microsoft.com/kb/934458
If SQL Server is between 3150 and 3158, get build 3159 from http://support.microsoft.com/kb/934459
Why everyone shouldn't just move to 3159, I'm not sure. I'm not quite ready to try it out yet, but they must be pretty confident this time, because they're pushing out through Windows Update as critical.
It's one of the more popular items on connect, unfortunately it is not a top-rated item because one party pooper voted 3. I think many people are afraid to vote because it looks like a slam on MS, but it's not – it's an appeal to make things right.
Aaron,
You should get the connect incident circulated so MS might take notice.
Wow, great list, I was searching for 9.00.315x, and I should have been searching for 2005.90.3156.0. Why different KB articles vary in the way they present the build/version number of sqlservr.exe, I will never understand.
I agree totally about SP3:
http://web.archive.org/web/*/http://web.archive.org/web/*/https://connect.microsoft.com/SQLServer/feedback/ViewFeedback.aspx?FeedbackID=267610
I have found this:- http://www.sqlservercentral.com/columnists/sjones/2960.asp
It mentions a couple of the builds.
Looking at this post why would MS issue a patch, marked as a critical update in Microsoft Update, that includes fixes that should not be used or even tested? The whole SP2 episode leaves a very sour taste. We had always gone with SP and security fix levels when we upgrade but now we have no real idea what we have gotten into. Based on this MS should issue a fully tested SP3 asap
Chris
I couldn't find anything public, so I think it's possible that these builds they speak of may only apply to customers who have obtained builds through PSS that aren't blessed enough for general distribution. In general, I think you will see that the fixes in 3152 are not included in 3042, and so 3159 would have those fixes as well, but 3054 would not. The bugs fixed in 3152 but not in 3040 are described here:
http://support.microsoft.com/?kbid=933097
I will try to find out more information but I can't promise anything.
I would like to know what fixes are in the builds between 3152 (the hotfix rollup we are testing with) and 3158 to have an idea what we may have to look out for if we want this new build 3159.
Chris