Re: SQL Server on Linux

On Wednesday, June 1, 2016 at 12:02:54 AM UTC+10, Mladen Gogala wrote:

> I used to like Pete Sharman, especially his work on oracle-l, but he
> seems to have switched to marketing and is now trying to sell OEM.

He's done some very good work there, at least the blessed thing became a bit more stable and usable. Until 13c. The jury on that one is still out, I'm afraid: I installed it with the latest patch set only to see the add node function crash with error ORA-1422 (unique fetch returns more than one row)! We rolled back the sub-patch that caused the problem and it's all working more or less now. It is completely inadmissible a "stability" patch ends up stopping us from adding nodes!!!

And haven't tested everything yet...

> 2) It doesn't have WebLogic. I don't use WebLogic, if it is up to me.

I hear you 5 on 5! :(

> That would probably put me at odds with Pete, had he not changed the
> profession.

Oh, I've been at odds with him for decades. But fortunately we always end up squaring it out over a quiet beer or two. :)

> useful for me as a techie on NYOUG. Oracle closed the technical
> information and is only dispatching the much needed technical info to
> chosen few, and not on the conferences.

Unbelievable! They tried that here and basically the whole community gave them the finger and stopped showing up at their functions. Simple as that!
This "chosen few" bullshit of their marketing MUST be forced to stop!

> Now that you mentioned 12c, it has
> multi-threaded log writer, based on polling, rather than post-wait event
> model. Of course, there is no documentation. The only documentation that
> I've found is how to turn it off, because of a bug which may cause the DB
> to hang.

So scary... And don't get me started on the "adaptive optimizer", which needs to be turned off if you install EM 13c! How's that for a "useful feature that everyone should be using"? And of course: if we don't, we are a "bad DBAs!". #facepalm

> decision was up to me. Oracle 10.1 was disastrously unstable, just like
> the early releases of 10.2. I wonder if you remember the bug in the
> 10.2.0.1 client which would hang after 200 days, necessitating a reboot?

Yegawds, do I ever! Was my first task when I joined here, to move everything off 10.1 as fast as I could push it!

> CPU cores. I have always considered parallel query to be reserved for
> very special queries, with the decision to use parallelism always
> warranting a careful consideration.

Indeed, agreed. I use auto degree of parallelism only for very small dbs with small number of users. Rarely if ever a problem there, provided I don't go feral with number of parallel server tasks. I've only used it once successfully in a larger db - by setting all tables to a fixed max degree of parallelism instead of DEFAULT. DEFAULT grabs as many parallel server tasks as it can 99.9999% of the time and bugger the effects on the rest of the load in the system...

> Bugs are to be expected, with such an extensive
> rewrite. And bugs and the lack of stability in the optimizer plans equal
> possible performance disaster in the critical applications.

Bingo!

> Personally, I
> would not switch before 12.2.0.3.

I'm finding 12.1.0.2.160119 borderline stable. But have already seen some alarms on the latest patch level after that! No end to the "stability" of 12c!...
Unreal! Received on Wed Jun 01 2016 - 06:45:16 CEST