Are Oracle DBAs trivialized?
Domenic
More and more I am finding that the role of the Oracle DBA is not
taken as seriously as it once was.
Having worked for many software companies in the last few years I see
the same pattern over and over -- most shops refuse to stay current
with Oracle and patchsets -- they're still running 8i, don't want to
use Oracle-specific features (clusters, IOTs, etc.) for fear it will
lock them into Oracle, and let the developers model and write their
own DDL.
Most of the time I feel they only want a DBA to put out fires.
Here are some recent examples ...
- I had to pull teeth to get my current company to switch to RMAN.
- I got laughed at for naming NOT NULL constraints in my CREATE TABLE
scripts
- I had a one of my designs tossed out because it had triggers,
clusters, and deferrable FK constraints in it.
- I have been telling management for almost a year to get off 8i --
falls on deaf ears of course
- I get apps tossed at me with hundreds of poorly written queries
(correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
rewrite for speed.
- I get crap from the UNIX SAs at just the thought using 10g ASM when
we finally upgrade, by that time Oracle 13 will be out.
- I have developers who commit every SQL statement and throw highly
denormalized tables at me, but management still feels they should NOT
have to come to me to get a table designed.
- HA requirements but they refuse to use RAC or Data Guard.
This is not just my current company, but they all seem to be the same
lately. It seems that the Java shops operate this way more than
anyone else -- very hostile attitude towards Oracle and DBAs in
general.
Is it just me, or does this go on everywhere?
Domenic.
premme...@hotmail.com
it is very difficult to use them in projects for reasons you mentioned.
Also, for last two years I have been working in the world of packaged
applications where all database design etc has been done by vendors.
Vendors want to run their application against many versions of Oracle
as well as against different databases such as SQL Server, DB2 UDB, so
they do not use anything new of Oracle.
DA Morgan
I'm going to take an admittedly contoversial approach to answering
your inquiry expecting others to throw a match at the gasoline I am
sloshing about. Here goes:
Most DBAs are totally incompetent and deserve little if any respect.
They were, past tense, competent when they learned their trade with
version 6.x or 7.x but they have not invested much effort in upgrading
their skills since then. They have self-defined their job as install,
manage users and schemas, patch, backup, and pray you don't have to
restore. They haven't read the Net Services book since it was 700 pages
which was a very long time, and more than 1000 new pages, ago.
They have not learned App Server.
They wouldn't know an ear file from a jar file.
They have not learned RMAN
They have not learned DataGuard
They have not learned DBMS_XPLAN
They have not even learned to read a StatsPack.
We are in the same position as were physicians before they started
policing their own ranks with certifications that had meaning and
requirements for continuing education to maintain certification. So
as our DBA community ages so does its skill set.
Now I didn't say all ... so keep your flames appropriate to the
provocation. ;-)
--
Daniel A. Morgan
University of Washington
damo...@x.washington.edu
(replace 'x' with 'u' to respond)
bdb...@gmail.com
> Most DBAs are totally incompetent and deserve little if any respect.
Well, isn't it everyone's dream in US employment to reach their level
of incompetence?
By the way, that is supposed to be a joke.
Obsolescence and outsourcing should scare them out of complacency.
You left off pl/sql coding.
Dataguard is only useful to those running EE.
I'd say that System Administration and Security would be areas worthy
of study, also but I doubt that Domenic (OP) could actually get any
security-related changes implemented.
-bdbafh
Tim X
> I'd like everyone's thoughts on this ...
>
> More and more I am finding that the role of the Oracle DBA is not
> taken as seriously as it once was.
>
> Having worked for many software companies in the last few years I see
> the same pattern over and over -- most shops refuse to stay current
> with Oracle and patchsets -- they're still running 8i, don't want to
> use Oracle-specific features (clusters, IOTs, etc.) for fear it will
> lock them into Oracle, and let the developers model and write their
> own DDL.
>
> Most of the time I feel they only want a DBA to put out fires.
>
> Here are some recent examples ...
>
> - I had to pull teeth to get my current company to switch to RMAN.
>
> - I got laughed at for naming NOT NULL constraints in my CREATE TABLE
> scripts
hehe - you see their thinking hasnt gone that far - they still don't
know which columns should and should not allow null (if any) because
no real analysis or design has occured.
>
> - I had a one of my designs tossed out because it had triggers,
> clusters, and deferrable FK constraints in it.
>
I suspect they wee tossed out because they didn't understand even what
a FK was.
> - I have been telling management for almost a year to get off 8i --
> falls on deaf ears of course
Still trying to get rid of an 8i db ourselves.
>
> - I get apps tossed at me with hundreds of poorly written queries
> (correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
> rewrite for speed.
>
We have actually started giving a can of 'spam' to the developer which
either does the largest cartesian join or the simplest select
statement with the record for consistent 'gets'.
> - I get crap from the UNIX SAs at just the thought using 10g ASM when
> we finally upgrade, by that time Oracle 13 will be out.
Lucky here - DBAs and Sys Admins get on well and both want to see us
get to 10g sooner rather than later.
>
> - I have developers who commit every SQL statement and throw highly
> denormalized tables at me, but management still feels they should NOT
> have to come to me to get a table designed.
We get that and the opposite - huge numbers of unrelated inserts
without any commits which keep blowing out rollback segments no matter
how large they are! Really seems to be know mental comprehension of
what a commit is and when and when not to use it.
>
> - HA requirements but they refuse to use RAC or Data Guard.
Same and it took nearly 12 months to convince the business and
developers of the merits of hot backups instead of nightly cold
backups!
>
> This is not just my current company, but they all seem to be the same
> lately. It seems that the Java shops operate this way more than
> anyone else -- very hostile attitude towards Oracle and DBAs in
> general.
Not sure if its 'hostile' or just plain old ignorance and a failure to
recognize the relevance of knowledge which is outside your own
specialist area (like java programming).
>
> Is it just me, or does this go on everywhere?
>
I'm no longer doing development or DBA work (I was only ever an Oracle
DBA 'in training' anyway). I made the fatal mistake of moving into
management and now have to manage the DBAs, Unix and Windows and
network sys admins.
We are very lucky in that we have a very experienced, capable and
professional senior DBA and a very capable and promising 'junior' DBA
(he isn't really junior, just not senior - if you get the
difference). We also have some very capable sys admins. Luckily, the
sys admins and DBAs have a very good and strong working
relationship. However, the DBAs do cop a lot of flak from some of the
developers and from some of the other managers. One of the managers
from the business side of things actually made the statement a few
weeks ago that DBAs were really just glorified helpdesk staff - you
see, he is an expert because he has a MSCE!
There is a constant battle with the developers because they feel the
DBAs (and sys admins) are simply there to make their life difficult -
they get upset when we reject work they have done because it is
inefficient, has security holes or is just a bad design.
Unfortunately, the style of management around these days is all about
short-term benefits and getting projects 'over the line'. Very little
is put into good solid long term solutions. I believe its because most
managers I've come across had no other real skill other than
'managing' and really only wanted to climb the career ladder. For
them, any solution which lasts more than 5 years is a waste as they
will be long gone onto that next promotion - they want quick short
term success at a low cost so that it looks like they have done their
job better and can get that promotion and fatter pay packet
sooner. therefore, they pay cheap inexperienced developers to get the
job done quick and dirty and don't want irritating DBAs getting in the
way.
I see all of the things you listed on nearly a daily basis. When I
first started in this industry, the DBA was a person you went to for
advice and guidence - but back then, nearly all the developers new at
least the first 3 normal forms, they new who Cobb was and beleived
they were building systems which would last and require minimal
maintenance.
The last laugh is with me though! I've been given the job of selecting
the new development platform for the developers. Currently, they only
use Oracle forms and reports. the problem is that we have changed
direction and no longer do really full blown development, but rather
integration of systems and setting up reporting frameworks etc. The
type of work now requires rapid turn-around and the Oracle tools just
don't cut it for this type of work. I've been given the job because
the developers couldn't agree on a direction.
All the developers are really worried as they know my favorite
language of all is Common Lisp. I've led them to believe this is what
I'm going to be putting into place - of course, there is no way I
could ever sell this to the business (unfortunately), but its fun watching the panic
and fear in their eyes!
In reality, I'll probably go with Java and a scripting language like
Perl (or python!).
Tim
--
Tim Cross
The e-mail address on this message is FALSE (obviously!). My real e-mail is
to a company in Australia called rapttech and my login is tcross - if you
really need to send mail, you should be able to work it out!
Tim X
> Most DBAs are totally incompetent and deserve little if any respect.
>
> They were, past tense, competent when they learned their trade with
> version 6.x or 7.x but they have not invested much effort in upgrading
> their skills since then. They have self-defined their job as install,
> manage users and schemas, patch, backup, and pray you don't have to
> restore. They haven't read the Net Services book since it was 700 pages
> which was a very long time, and more than 1000 new pages, ago.
>
> They have not learned App Server.
> They wouldn't know an ear file from a jar file.
> They have not learned RMAN
> They have not learned DataGuard
> They have not learned DBMS_XPLAN
> They have not even learned to read a StatsPack.
>
> We are in the same position as were physicians before they started
> policing their own ranks with certifications that had meaning and
> requirements for continuing education to maintain certification. So
> as our DBA community ages so does its skill set.
>
> Now I didn't say all ... so keep your flames appropriate to the
> provocation. ;-)
Maybe I've been even luckier than I thought - the DBAs I've worked
with have been very good. However, I would point out some mitigation
for your observations - I'd suggest a big part of the problem here is
the lack of time made available for DBAs to keep their knowledge up to
date. while I do beleive every professional ha a responsability to
maintain their specialist knowledge and a good part of it should be
maintained in their own time, I also feel employers should be
encouraging and supporting this process. Most of the DBAs I've worked
with have been required to do large amounts of 'out of hours' work as
well as their normal hours work. Ive had to fight and argue very hard
to get the funds necessary to send our DBAs on training and to
relevant conferences.
The drop in skill amongst DBAs is probably paralleled with the drop in
skill of developers generally. I've seen very little investment put
into either area and have noticed a declining wage and increasing
expectations for hours of work in both areas. Business/employes need
to invest in these resources if they want long term benefits - but
most don't.
Noons
> I'm going to take an admittedly contoversial approach to answering
> your inquiry expecting others to throw a match at the gasoline I am
> sloshing about. Here goes:
>
> Most DBAs are totally incompetent and deserve little if any respect.
I'm inclined to agree with you, for essentially the same
reasons. The number of "pretend" dba's out there is
staggering. The only difference in our points of view
is that I see it happening with dba's of ALL ages, not just
the old ones...
--
Cheers
Nuno Souto
in sunny Sydney, Australia
wizo...@yahoo.com.au.nospam
Mark Bole
> dome...@hotmail.com (Domenic) writes:
>
>
>>I'd like everyone's thoughts on this ...
>>
>>More and more I am finding that the role of the Oracle DBA is not
>>taken as seriously as it once was.
>>
[...]
> Unfortunately, the style of management around these days is all about
> short-term benefits and getting projects 'over the line'. Very little
> is put into good solid long term solutions. I believe its because most
> managers I've come across had no other real skill other than
> 'managing' and really only wanted to climb the career ladder. For
> them, any solution which lasts more than 5 years is a waste as they
> will be long gone onto that next promotion - they want quick short
> term success at a low cost so that it looks like they have done their
> job better and can get that promotion and fatter pay packet
> sooner. therefore, they pay cheap inexperienced developers to get the
> job done quick and dirty and don't want irritating DBAs getting in the
> way.
[...]
>
> Tim
>
You beat me to the punch, being the first respondent in this thread to
identify poor management as the underlying cause of the problem, instead
of just relating the symptoms (isn't that what good DBA's are supposed
to do? ;-)
-Mark Bole
DA Morgan
So is partitioning option. But any DBA that doesn't have EE installed
on a machine at home, or a laptop so that they can study the technology
is not doing their job. To return to my physician analogy ... you don't
learn about neurosurgery after the patient is on the gurney.
DA Morgan
I agree with all of your points.
DBAs, and developers too, should insist on the cost of training being
included with the cost of any upgrade planning.
DA Morgan
> DA Morgan apparently said,on my timestamp of 11/04/2005 2:15 PM:
>
>> I'm going to take an admittedly contoversial approach to answering
>> your inquiry expecting others to throw a match at the gasoline I am
>> sloshing about. Here goes:
>>
>> Most DBAs are totally incompetent and deserve little if any respect.
>
>
> I'm inclined to agree with you, for essentially the same
> reasons. The number of "pretend" dba's out there is
> staggering. The only difference in our points of view
> is that I see it happening with dba's of ALL ages, not just
> the old ones...
Alas I agree with you too.
No matches tossed into the petrol. I'm losing my touch. ;-)
xho...@gmail.com
> I'd like everyone's thoughts on this ...
>
> More and more I am finding that the role of the Oracle DBA is not
> taken as seriously as it once was.
>
> Having worked for many software companies in the last few years I see
> the same pattern over and over -- most shops refuse to stay current
> with Oracle and patchsets -- they're still running 8i, don't want to
> use Oracle-specific features (clusters, IOTs, etc.) for fear it will
> lock them into Oracle, and let the developers model and write their
> own DDL.
When I waited for the DBA to model and write the DDL, it simply never
got done. I was forced, as a developer (well, not even that, more
like as an advanced end user, really) to do the modelling and DDL myself
if I actually wanted to have a schema. And then when I gave the DDL to the
then DBA to implement (along with expected growth rates), he complained it
didn't have storage clauses. You are the DBA, how about you add the
storage clauses?
>
> Most of the time I feel they only want a DBA to put out fires.
>
> Here are some recent examples ...
Most of these things don't sound specific to only DBAs at all, it is merely
an industry-wide Dilbert cartoon, viewed through the eyes of a DBA.
>
> - I had to pull teeth to get my current company to switch to RMAN.
>
> - I got laughed at for naming NOT NULL constraints in my CREATE TABLE
> scripts
Well, this one I kind of see. Why would a numeric not null column need a
separately named not null constraint any more than it needs a separately
named numericity constraint. Both of these constraints are inherent to the
column definition, are they not?
>
> - I had a one of my designs tossed out because it had triggers,
> clusters, and deferrable FK constraints in it.
>
> - I have been telling management for almost a year to get off 8i --
> falls on deaf ears of course
>
> - I get apps tossed at me with hundreds of poorly written queries
> (correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
> rewrite for speed.
Well, isn't that part of your job? Is your complaint that the developers
aren't tuning their own queries (which if they did would make DBAs even
more trivial) or that developers are even writing queries in the first
place?
> - I get crap from the UNIX SAs at just the thought using 10g ASM when
> we finally upgrade, by that time Oracle 13 will be out.
>
> - I have developers who commit every SQL statement and throw highly
> denormalized tables at me, but management still feels they should NOT
> have to come to me to get a table designed.
>
> - HA requirements but they refuse to use RAC or Data Guard.
Are they legitimate HA requirements, or does someone just like saying
24/7 because it makes them sound hip?
> This is not just my current company, but they all seem to be the same
> lately. It seems that the Java shops operate this way more than
> anyone else -- very hostile attitude towards Oracle and DBAs in
> general.
>
> Is it just me, or does this go on everywhere?
Not just everywhere, but with every other profession as well.
Xho
--
-------------------- http://NewsReader.Com/ --------------------
Usenet Newsgroup Service $9.95/Month 30GB
BigBo...@hotmail.com
> (correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
> rewrite for speed.
Just curious, why does the presence of correlated subqueries or NOT
EXISTS clauses equate to a poorly written query? As a DBA & developer,
I've used them quite frequently and they perform just fine. Is there a
more efficient alternative?
-Steve
DA Morgan
No. Certainly they are not the silver bullet solution usable in ALL
situations ... but no.
Sybrand Bakker
>dome...@hotmail.com (Domenic) wrote:
>> I'd like everyone's thoughts on this ...
>>
>> More and more I am finding that the role of the Oracle DBA is not
>> taken as seriously as it once was.
>>
>> Having worked for many software companies in the last few years I see
>> the same pattern over and over -- most shops refuse to stay current
>> with Oracle and patchsets -- they're still running 8i, don't want to
>> use Oracle-specific features (clusters, IOTs, etc.) for fear it will
>> lock them into Oracle, and let the developers model and write their
>> own DDL.
>
>When I waited for the DBA to model and write the DDL, it simply never
>got done. I was forced, as a developer (well, not even that, more
>like as an advanced end user, really) to do the modelling and DDL myself
>if I actually wanted to have a schema. And then when I gave the DDL to the
>then DBA to implement (along with expected growth rates), he complained it
>didn't have storage clauses. You are the DBA, how about you add the
>storage clauses?
My experience is that most developers don't know how to set up a
datamodel and have never heard of normalization.
>
>>
>> Most of the time I feel they only want a DBA to put out fires.
>>
>> Here are some recent examples ...
>
>Most of these things don't sound specific to only DBAs at all, it is merely
>an industry-wide Dilbert cartoon, viewed through the eyes of a DBA.
>
>>
>> - I had to pull teeth to get my current company to switch to RMAN.
>>
>> - I got laughed at for naming NOT NULL constraints in my CREATE TABLE
>> scripts
>
>Well, this one I kind of see. Why would a numeric not null column need a
>separately named not null constraint any more than it needs a separately
>named numericity constraint. Both of these constraints are inherent to the
>column definition, are they not?
>
>>
>> - I had a one of my designs tossed out because it had triggers,
>> clusters, and deferrable FK constraints in it.
>>
>> - I have been telling management for almost a year to get off 8i --
>> falls on deaf ears of course
>>
>> - I get apps tossed at me with hundreds of poorly written queries
>> (correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
>> rewrite for speed.
>
>Well, isn't that part of your job? Is your complaint that the developers
>aren't tuning their own queries (which if they did would make DBAs even
>more trivial) or that developers are even writing queries in the first
>place?
>
His complaint is that most developers don't KNOW how to tune their
queries. Consequently the crapplication work as long as it contains 50
records. As soon as it gets into production the crapplication
collapses and developers start blaming the DBA, who 'doesn't know
anything about databases'
>
>> - I get crap from the UNIX SAs at just the thought using 10g ASM when
>> we finally upgrade, by that time Oracle 13 will be out.
>>
>> - I have developers who commit every SQL statement and throw highly
>> denormalized tables at me, but management still feels they should NOT
>> have to come to me to get a table designed.
>>
>> - HA requirements but they refuse to use RAC or Data Guard.
>
>Are they legitimate HA requirements, or does someone just like saying
>24/7 because it makes them sound hip?
>
>> This is not just my current company, but they all seem to be the same
>> lately. It seems that the Java shops operate this way more than
>> anyone else -- very hostile attitude towards Oracle and DBAs in
>> general.
>>
>> Is it just me, or does this go on everywhere?
>
>Not just everywhere, but with every other profession as well.
>
>Xho
--
Sybrand Bakker, Senior Oracle DBA
Joel Garry
>on a machine at home, or a laptop so that they can study the
technology
>is not doing their job. To return to my physician analogy ... you
don't
>learn about neurosurgery after the patient is on the gurney.
I am now picturing Dr. Frankenstein practicing his neurosurgery at
home.
I don't have time to go "shopping" (ie checking wrong linux
compatibility lists, taking incorrect firewire advice, and on anon, not
to mention actually _buying_ one) for a PC that could run EE because
I'm too busy working on it at work.
jg
--
@home.com is bogus. "It takes intelligence, concentration, process and
detail orientation, anal retentiveness, and other such endearing
personality traits to make a good dba. Those traits are AS IMPORTANT as
technical skills." - Michael from Allentown, PA
Tim X
>
> So is partitioning option. But any DBA that doesn't have EE installed
> on a machine at home, or a laptop so that they can study the technology
> is not doing their job. To return to my physician analogy ... you don't
> learn about neurosurgery after the patient is on the gurney.
I think this is quite a valid point. Many years ago, it was difficult
to get a platform you could run at home and 'play' with to learn your
trade - desktop machines were not up to it and often you couldn't get
a version which would run on something an individual could afford.
However, these days, therre is just no excuse - between windows and
linux, there is little you cannot run at home on a PC - even so called
'enterprise editions'.
Personally, I just don't understand the resistance some have to
learning 'out of hours' - while I argued earlier in this thread that
management needs to invest more in the training etc, I still strongly
believe that as a professional, you have a responsability to keep 'up
to speed' yourself - if for no other reason thatn to make sure you
keep your employment options open.
When working as a developer and when working in the DBA area, I always
made sure I had an environment at home which was as far as possible a
mirror of my work environment - I would use this as a place to learn,
experiment and try out new ideas.
Lewis C
>When I waited for the DBA to model and write the DDL, it simply never
>got done. I was forced, as a developer (well, not even that, more
>like as an advanced end user, really) to do the modelling and DDL myself
>if I actually wanted to have a schema. And then when I gave the DDL to the
>then DBA to implement (along with expected growth rates), he complained it
>didn't have storage clauses. You are the DBA, how about you add the
>storage clauses?
>
That happens. If you developed the DDL as a compromise (becasue they
didn't have the time), the DBAs should have made their expectations
known and let you know in advance what kind of storage they were
looking for.
>> - I got laughed at for naming NOT NULL constraints in my CREATE TABLE
>> scripts
>
>Well, this one I kind of see. Why would a numeric not null column need a
>separately named not null constraint any more than it needs a separately
>named numericity constraint. Both of these constraints are inherent to the
>column definition, are they not?
>
Export and import mostly. Named constraints are good. If you have an
unamed not null constraint, export the schema, import to a different
database with an unnamed constraint of the same name, kerblooey. If a
constraint is not explicitly named, oracle will name it for you. SQL
standards would help there as would using a design tool instead of
hand coding everything.
>> - I get apps tossed at me with hundreds of poorly written queries
>> (correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
>> rewrite for speed.
>
>Well, isn't that part of your job? Is your complaint that the developers
>aren't tuning their own queries (which if they did would make DBAs even
>more trivial) or that developers are even writing queries in the first
>place?
That really depends on the environment. I've been places where
developers were expected to tune things themselves and places where
the DBAs had to babysit. If the developers have the skills, or want
to learn them, that should be supported by the DBAs. If they don't
have the skills, the DBAs should communicate to their management the
problem of tuning complete applications after they've been built.
This one would probably require a management change.
>>
>> Is it just me, or does this go on everywhere?
>
>Not just everywhere, but with every other profession as well.
>
Yep. I agree with that.
Lewis
-----------------------------------------------------------
Lewis R Cunningham
Author, ItToolBox Blog: An Expert's Guide to Oracle
http://blogs.ittoolbox.com/oracle/guide/
Topic Editor, Suite101.com: Oracle Database
http://www.suite101.com/welcome.cfm/oracle
Sign up for courses here:
http://www.suite101.com/suiteu/default.cfm/416752
-----------------------------------------------------------
HF
training and ensure i have enough time in a day to learn. It is an
expectation that all dba's senior and junior thake the time to do at least
an hour a day of reserach/training to keep them up to date and continually
improving their skills. Is it the fault of an overworked DBA who doesnt get
the time in work hours for training so that they can do their job better.
Why should it be an expectation that a DBA must do addition hours at home so
that he can do his job better for his disgruntle employer, who makes no
effort in looking after his employees. At what point did work become part of
the home life?
offsho...@hotmail.com
home. To learn new skills and keep up with latest technologies one
has to find a work environment. The root cause of the problem at least
in USA is the way industries operate. Job's have been made so narrowly
specialized. One is forced to do same thing over and over again. Why
one cannot do both development and DBA work especially when a person
knows both. I could never find a job like that. I always had to take
off my development skills from resume to find a DBA job.
Industries are constantly eliminating jobs, hire only temporary
workers and keep sending jobs offshore. They only want people with
many years of experience in some narrow area. In last three years, I
am forced to change companies every year. As soon as I arrive in a
company, I find that people from offshore are coming for training and
majority of IT work will be taken over by them. Can someone tell me in
a work environment like that how you can prosper?
So name of the game is to pray and hope you can keep finding some or
other job until you are in your grave. Executives in corporations in
USA (may be true of other countries) are ugly, self centered, with no
interest in employees or long term interest of country. Only thing
which matters is how to get rich quickly.
DA Morgan is fortunate, he has a University Professor job where he has
opportunity to try different things, learn new skills, etc. He does
not have to worry every day about some corporate merger, management
change, or disappearance of his job. So, he is probably thinking that
everyone has same opportunity and all DBA's should be at his level.
Let us face it, the way industries operate, there is not that much one
can do in terms of improving skill set. If no one gives you a job as a
developer, you can learn all the programming languages in your spare
time and sooner or later you will loss interest in them and your skills
will become outdated. A friend of mine who did Ph.D in Computer
Science, has many years of experience in software development, was let
go because IBM sent this job to India. He could not find a developer or
any other IT job in Industry. He was forced to become a substitute
teacher in his township public school. He is not alone, in USA number
of people like him keep increasing. Does any one have idea what one can
do in terms of skillset improvement to stop that. I don't think you
can do anything.
D Rolfe
Disclaimer: I work for a company that makes a product that
writes Java to run run SQL and PL/SQL statements...
-------------------------------------------------------------
Domenic wrote:
> I'd like everyone's thoughts on this ...
>
> More and more I am finding that the role of the Oracle DBA is not
> taken as seriously as it once was.
>
> Having worked for many software companies in the last few years I see
> the same pattern over and over -- most shops refuse to stay current
> with Oracle and patchsets -- they're still running 8i, don't want to
> use Oracle-specific features (clusters, IOTs, etc.) for fear it will
> lock them into Oracle, and let the developers model and write their
> own DDL.
>
> Most of the time I feel they only want a DBA to put out fires.
>
> Here are some recent examples ...
>
...
>
> This is not just my current company, but they all seem to be the same
> lately. It seems that the Java shops operate this way more than
> anyone else -- very hostile attitude towards Oracle and DBAs in
> general.
>
I think there is a general pattern of new computer science graduates not
taking databases as seriously as they used to. DBAs are the
anesthesiologists of the software industry - unfashionable, but try
building a real system without them!
I have first hand experience of developers who regard the database as a
'storage device'. I had to spend *weeks* fighting with a developer on
his second job out of college who's idea of using a database for on-line
mortgage applications was to serialize his Java objects and store them
in BLOB columns. Can you imagine what writing the reports would have
been like?
Unlike COBOL, C or C++, modeling an entire application in a computer
language as opposed to the database is not only possible but practical
with Java, with the result being that the people who do the actual
coding don't see the value in having business logic and rules at the
database level. Current notions about multi-tier architectures also come
into this, with some developers regarding things like foreign key
constraints as 'business logic' that should be dealt with at the
'application server' layer.
When it comes to Java and Oracle the JDBC standard is also a factor. The
capabilities of PL/SQL are so far ahead of what JDBC will easily support
that it's very easy for a PL/SQL developer to write a procedure that
can't be run from Java without the user of a pre-compiler, wrapper
procedures or laboriously hand-coding anonymous blocks (this is the
issue my employer addresses). If the DBA's keep producing procedures
that the Java developers can't use without jumping through hoops a
degree of frustration and even hostility is hardly surprising.
Oracle is also a victim of its own success when it comes to 8i. 8i is
good. So good that there's no *compelling* incentive to upgrade to 9i or
10g. People in the Oracle community know that an upgrade will have to be
done sooner or later and that the new versions are indeed better but in
a lot of cases management will perceive it as more money to solve a
non-existent problem.
My 10 cents..
David Rolfe
Orinda Software
Dublin, Ireland
www.orindasoft.com
D Rolfe
Disclaimer: I work for a company that makes a product that
writes Java to run run SQL and PL/SQL statements...
-------------------------------------------------------------
Domenic wrote:
> I'd like everyone's thoughts on this ...
>
> More and more I am finding that the role of the Oracle DBA is not
> taken as seriously as it once was.
>
> Having worked for many software companies in the last few years I see
> the same pattern over and over -- most shops refuse to stay current
> with Oracle and patchsets -- they're still running 8i, don't want to
> use Oracle-specific features (clusters, IOTs, etc.) for fear it will
> lock them into Oracle, and let the developers model and write their
> own DDL.
>
> Most of the time I feel they only want a DBA to put out fires.
>
> Here are some recent examples ...
>
...
>
> This is not just my current company, but they all seem to be the same
> lately. It seems that the Java shops operate this way more than
> anyone else -- very hostile attitude towards Oracle and DBAs in
> general.
>
I think there is a general pattern of new computer science graduates not
taking databases as seriously as they used to. DBAs are the
anesthesiologists of the software industry - unfashionable, but try
building a real system without them!
I have first hand experience of developers who regard the database as a
'storage device'. I had to spend *weeks* fighting with a developer on
his second job out of college who's idea of using a database for on-line
mortage applications was to serialize his Java objects and store them in
BLOB columns. Can you imagine what writing the reports would have been like?
Unlike COBOL, C or C++, modeling an entire application in a computer
language as opposed to the database is not only possible but practical
with Java, with the result being that the people who do the actual
coding don't see the value in having business logic and rules at the
database level. Current notions about multi-tier architechtures also
come into this, with some developers regarding things like foreign key
constraints as 'business logic' that should be dealt with at the
'application server' layer.
When it comes to Java and Oracle the JDBC standard is also a factor. The
capabilities of PL/SQL are so far ahead of what JDBC will easily support
that it's very easy for a PL/SQL developer to write a procedure that
can't be run from Java without the user of a pre-compiler, wrapper
procedures or laboriously hand-coding anonymous blocks (this is the
issue my employer addresses). If the DBA's keep producing procedures
that the Java developers can't use without jumping through hoops a
degree of frustration and even hostility is hardly surprising.
Oracle is also a victim of its own success when it comes to 8i. 8i is
good. So good that there's no *compelling* incentive to upgrade to 9i or
10g. People in the Oracle community know that an upgrade will have to be
done sooner or later and that the new versions are indeed better but in
a lot of cases management will perceive it as more money to solve a
non-existant problem.
Frank van Bortel
> DA Morgan <damo...@x.washington.edu> writes:
>
>
>>So is partitioning option. But any DBA that doesn't have EE installed
>>on a machine at home, or a laptop so that they can study the technology
>>is not doing their job. To return to my physician analogy ... you don't
>>learn about neurosurgery after the patient is on the gurney.
>
>
> I think this is quite a valid point. Many years ago, it was difficult
> to get a platform you could run at home and 'play' with to learn your
> trade - desktop machines were not up to it and often you couldn't get
> a version which would run on something an individual could afford.
>
Many, many, MANY years, then... I have been running Oracle on
laptops (and desktops) for about 10 years...
> However, these days, therre is just no excuse - between windows and
> linux, there is little you cannot run at home on a PC - even so called
> 'enterprise editions'.
Explain the differences in options between Oracle Personal, and
Oracle Enterprise editions.
For learning purposes, none.
As for 'footprint' - that's up to you: you may not need to copy
all your production system parameter settings.
>
> Personally, I just don't understand the resistance some have to
> learning 'out of hours' - while I argued earlier in this thread that
> management needs to invest more in the training etc, I still strongly
> believe that as a professional, you have a responsability to keep 'up
> to speed' yourself - if for no other reason thatn to make sure you
> keep your employment options open.
Agree, but I feel there should be some reward in this off-hour study;
allow to visit (E)OUG meetings during office hours, all expenses paid.
That kind of stuff.
>
> When working as a developer and when working in the DBA area, I always
> made sure I had an environment at home which was as far as possible a
> mirror of my work environment - I would use this as a place to learn,
> experiment and try out new ideas.
>
Now, with OAS, that may be over the top for some desktops. :)
--
Regards,
Frank van Bortel
Tim X
>
> Agree, but I feel there should be some reward in this off-hour study;
> allow to visit (E)OUG meetings during office hours, all expenses paid.
> That kind of stuff.
>
Agreed. I think there should always be more carrot than stick!
>
> Now, with OAS, that may be over the top for some desktops. :)
>
Yes - very true - its all getting very large an complex and you
probably do need to select what is needed rather than everything -
though disk is getting cheapre and cpus are getting faster! My main
point is that you can now have an environment on your laptop which is
increadibly close to your production environment and really be able to
use good old 'suck and see' approaches to learning and be very
confident the results you observe are likely to be extremely close to
what you would see on a full blown production system.
Maybe I'm a little thick sometimes, but I have always found the
ability to try out my understanding and clarify my knowledge by
testing it invaluable. I often find descriptions and manuals lacking
in crucial bits of knowledge or not quite being clear enough for me to
grasp - the ability to freely experiment in an environment close to
the production system. There have been times in the past when I've
tried something out on a 'stripped down' version or on a really
different platform (windows v unix like systems etc) and thought I had
a grasp on it all only to find there was some subtle but critical
difference between the two - not saying this is the case with Oracle
personal edition or windows edition, but it is something which has
happened with other systems.
Tim X
Work has been part of the 'home life' for a long long time. I think
that if you want to reach the same level of 'professional'
clasification as medical doctors/specialists, barristers/lawyers,
architects etc, then there is an expectation of some personal
professional development. If on the other hand you want to look at
professionalism in the vain of plumbers, carpenters, middle managers
etc, then there may be less requirement to do as much 'at home'
learning - but then again, the rate of change in these professions
tends to be a lot slower than in IT. However, what really frustrates
me is the fact too many employers seem to want professionals in the
vain of Barristers and doctors, but only want to pay the salary of
middle managers or less!
I guess it all depends on what you want to aim at and what really
matters to you. I do feel there is too much of a trend towards longer
work hours and too much encroachment on our personal lives by our
professional lives and employers are all too often expecting too much
- I guess its all about what the market will bare!
DA Morgan
HF wrote:
> I am a new DBA and still in training, however my company pays for all my
> training and ensure i have enough time in a day to learn.
You are lucky and your situation is not the norm. Relish it.
> Why should it be an expectation that a DBA must do addition hours at home so
> that he can do his job better for his disgruntle employer, who makes no
> effort in looking after his employees. At what point did work become part of
> the home life?
One should learn at home not for the employer but for oneself. How are
you going to become proficient with RMAN if your employer doesn't use
it? Are you going to fake your next job interview? How about DataGuard?
How about RAC? How about any one of a huge number of options.
This is not about your employer ... it is about yourself.
And it should be noted that physicians, attorneys, engineers, CPAs, etc.
have been doing it for decades.
Paul
DA Morgan <damo...@x.washington.edu> wrote:
>> Just curious, why does the presence of correlated subqueries or NOT
>> EXISTS clauses equate to a poorly written query? As a DBA & developer,
>> I've used them quite frequently and they perform just fine. Is there a
>> more efficient alternative?
>No. Certainly they are not the silver bullet solution usable in ALL
>situations ... but no.
But haven't you been told that partitioning tables is a "silver
bullet"? 8-)
Paul...
--
plinehan __at__ yahoo __dot__ __com__
XP Pro, SP 2,
Oracle, 9.2.0.1.0 (Enterprise Ed.)
Interbase 6.0.2.0;
When asking database related questions, please give other posters
some clues, like operating system and version of db being used.
The exact text and/or number of error messages is useful (!= "it didn't work!").
Thanks.
Furthermore, As a courtesy to those who spend
time analyzing and attempting to help, please
do not top post.
Paul
Sybrand Bakker <pos...@sybrandb.demon.nl> wrote:
> Consequently the crapplication work as long as it contains 50
"Crapplication" - now there's a word that's going into my internet
dictionary.
Paul...
(puzzled by some of his so-called "DBA's".
HF
> You are lucky and your situation is not the norm. Relish it.
It isnt the norm because people are willing to let their working life
infultrate their home life. It should be the norm that employees give you
the training you need to do your job better for them.
> One should learn at home not for the employer but for oneself. How are
> you going to become proficient with RMAN if your employer doesn't use
> it? Are you going to fake your next job interview? How about DataGuard?
> How about RAC? How about any one of a huge number of options.
These are technologies that need to be learn to determine if they will suit
their business needs. How can your employers determine if they need these
technologies if none of their workers know anything about it. Our company
does not use RAC but 2 of us have been trained and learnt about this in work
hours to determine if our company would benifit from it. The company saves
so much in the long run. Who wants to spend millions on a technology that
they dont need (thanks to fast talking sales people), with no staff with the
knowledge to use it. Up until 6 months ago, we were not using RMAN because
none of the workers really knew anything about it. Someone got the training,
relised the benifits and now we are have all had training in it and use it
on a day to day basis.
Joel Garry
Tim X wrote:
> DA Morgan <damo...@x.washington.edu> writes:
>
> >
> > So is partitioning option. But any DBA that doesn't have EE
installed
> > on a machine at home, or a laptop so that they can study the
technology
> > is not doing their job. To return to my physician analogy ... you
don't
> > learn about neurosurgery after the patient is on the gurney.
>
> I think this is quite a valid point. Many years ago, it was difficult
> to get a platform you could run at home and 'play' with to learn your
> trade - desktop machines were not up to it and often you couldn't get
> a version which would run on something an individual could afford.
How many years are you talking? In the 80's, most of the _serious_
people I knew were buying non-pc platforms. I bought my first PDP in
'83. Through the decades, I've seen people buy Vaxen, Sun and so
forth, for personal pleasure and/or growth. It seems to be pretty
constant at around $3K (or less, for the lucky trash-diggers). Even
now, if you include some decent backup on a PC, it will cost near that
much - and how many people buy networks and multiple computers these
days?
>
> However, these days, therre is just no excuse - between windows and
> linux, there is little you cannot run at home on a PC - even so
called
> 'enterprise editions'.
>
> Personally, I just don't understand the resistance some have to
> learning 'out of hours' - while I argued earlier in this thread that
> management needs to invest more in the training etc, I still strongly
> believe that as a professional, you have a responsability to keep 'up
> to speed' yourself - if for no other reason thatn to make sure you
> keep your employment options open.
It implies worker abuse. How many developers have you known in a
60-100hr/wk death march - on salary? Companies that don't allow for
work-time growth and education, time off, pay for actual hours worked,
are rewarded with "increased productivity." And that's with some laws
against such abuse! Imagine what an unfettered (read: "slave") market
would do. Maybe it's different in your area.
>
> When working as a developer and when working in the DBA area, I
always
> made sure I had an environment at home which was as far as possible a
> mirror of my work environment - I would use this as a place to learn,
> experiment and try out new ideas.
Gave up on that a long time ago. I figure I've saved $1500/year simply
by becoming a cheap curmudgeon and getting cable. I also don't want to
compete with high school kids or offshoring. If the database isn't
big/concurrent/specialized enough to not work in some kiddiedb, I don't
have much interest in it. And if it is that specialized, it probably
has some licensing issues that won't allow running at home. (I admit
it, I want a cool new computer, parts of my frankenpc date to the
'80s).
jg
--
@home.com is bogus. "What the hell did you have to say that for?
You're in no position to claim 100 percent effectiveness. What's the
matter with you?" - Dr. Thomas Francis, scolding Jonas Salk after the
announcement of the polio vaccine, April 12, 1955
DA Morgan
> DA Morgan <damo...@x.washington.edu> wrote:
>
>>No. Certainly they are not the silver bullet solution usable in ALL
>>situations ... but no.
>
> But haven't you been told that partitioning tables is a "silver
> bullet"? 8-)
>
> Paul...
Not by Oracle. Though since it is Q4 I'm sure I can a some sales
person short of quota that would tell me anything I wanted to hear.
Shabble
suit
> their business needs. How can your employers determine if they need these
> technologies if none of their workers know anything about it.
>
>
The world is not yet ready for computerisation; because the statement above
is so true.
But we have computers I hear you say, plenty of them, all over the place!
Yes you have, but is your business computerised?
No, it isn't.
You have lots of lovely graphic kits, ERP systems, text processors,
spreadsheets, and databases, control systems etc;
but very little "real" integration. You are talking about the DBA job being
trivialised, yes I have seen and endured it myself.
I can assure you however, this is a temporary political phenomena.
Don't underestimate yourselves; you have the technical grounding to
understand the application
the new waves of technology, computerisation, which is coming.
Here's an example in terms you are already experts in.
Stand back for a minute and think about the differences between Oracle7
(Oracle6 if your my age) and compare it to Oracle10G.
Then, factor in the short period of time between the technologies. Pretty
stunning huh!
Haven't Oracle done an amazing job. I use the word amazing deliberately.
Look up the word "amaze" in a dictionary :-To affect with great wonder; to
astonish.
Yes, and the bye the way it also means: To bewilder; perplex.
Let the politicians or managers have their way, for the moment, the future
of computing is computerisation -
bringing it all together, were not there yet, but were on the way.
Who are going to be the amazed and who are going to be the perplexed, it's
up to you.
Shabs.
Jim Smith
<dome...@hotmail.com> writes
>I'd like everyone's thoughts on this ...
>
>More and more I am finding that the role of the Oracle DBA is not
>taken as seriously as it once was.
>
>Having worked for many software companies in the last few years I see
>the same pattern over and over -- most shops refuse to stay current
>with Oracle and patchsets -- they're still running 8i, don't want to
>use Oracle-specific features (clusters, IOTs, etc.) for fear it will
>lock them into Oracle, and let the developers model and write their
>own DDL.
>
There are two different issues here - development methods/styles and use
of Oracle features, but the both ultimately come down to cost/benefit.
The knack is not to argue in technical terms, but to point out the
benefits and costs. Admittedly this is not always straightforward.
>Most of the time I feel they only want a DBA to put out fires.
>
>Here are some recent examples ...
>
>- I had to pull teeth to get my current company to switch to RMAN.
Why should they switch to RMAN? The currently backups presumably work
fine, and there will be a significant cost in terms of DBA (and other
time) in making the switch and testing it.
Do the benefits of RMAN outweigh these costs (and the risks associated
with the disruption of something as fundamental as backups?) Almost
certainly, but some figures in terms of DBA time saved, shorter backup
windows, faster restores etc, would help the argument.
>
>- I got laughed at for naming NOT NULL constraints in my CREATE TABLE
>scripts
>
I'm afraid I would laugh too. Life's too short.
>- I had a one of my designs tossed out because it had triggers,
>clusters, and deferrable FK constraints in it.
>
Why? Non-portability? Performance concerns? Not making your developers
use these things is one thing, actively rejecting them is another.
>- I have been telling management for almost a year to get off 8i --
>falls on deaf ears of course
>
This is tricky. There are huge costs to upgrading, but the risks of
hitting a bug for which there is no fix are unquantifiable. In terms of
the benefits of later versions, again they need to be quantified. If the
shop is just using the database as a persistence store, then benefits
other than performance improvements are hard to see.
>- I get apps tossed at me with hundreds of poorly written queries
>(correlated subqueries everywhere, NOT EXISTS, etc.) that I have to
>rewrite for speed.
>
This ought just to be part of the performance testing of the app anyway,
whether you do the re-writing or the developers. (I accept that
performance testing probably isn't done but one can dream.)
>- I get crap from the UNIX SAs at just the thought using 10g ASM when
>we finally upgrade, by that time Oracle 13 will be out.
>
This is just an old fashioned turf war. (See the thread (Are UNIX SAs
trivialised?)[1]. But again, there are probably benefits in terms of
their resources which could be pointed out.
>- I have developers who commit every SQL statement and throw highly
>denormalized tables at me, but management still feels they should NOT
>have to come to me to get a table designed.
>
>- HA requirements but they refuse to use RAC or Data Guard.
>
>This is not just my current company, but they all seem to be the same
>lately. It seems that the Java shops operate this way more than
>anyone else -- very hostile attitude towards Oracle and DBAs in
>general.
>
>Is it just me, or does this go on everywhere?
>
>Domenic.
--
Jim Smith
Because of their persistent net abuse, I ignore mail from
these domains (among others) .yahoo.com .hotmail.com .kr .cn .tw
For an explanation see <http://www.jimsmith.demon.co.uk/spam>
Sybrand Bakker
<j...@jimsmith.demon.co.uk> wrote:
>
>Why should they switch to RMAN? The currently backups presumably work
>fine, and there will be a significant cost in terms of DBA (and other
>time) in making the switch and testing it.
>
>Do the benefits of RMAN outweigh these costs (and the risks associated
>with the disruption of something as fundamental as backups?) Almost
>certainly, but some figures in terms of DBA time saved, s
Apparently, you don't know RMAN, otherwise you wouldn't have made
those comments.
The company still NOT using Rman is not be envied, as Oracle clearly
considers RMAN to be core technology, interweaving it more and more
with the core database product.
As for your remarks on not upgrading: do you recommend companies to
dump their support contract if they don't plan to upgrade.? Probably
you don't. But if they don't upgrade, what is the value of their
support contract?
All in all I get the feeling you will soon get trivialized, as you
don't seem to realize once you stop keeping your skills current,
inevitably *you* will be dumped sometime, and stay unemployed, as your
cv qualifies you as a dinosaur.
Jim Smith
<pos...@sybrandb.demon.nl> writes
>On Sun, 17 Apr 2005 11:54:59 +0100, Jim Smith
><j...@jimsmith.demon.co.uk> wrote:
>
>>
>>Why should they switch to RMAN? The currently backups presumably work
>>fine, and there will be a significant cost in terms of DBA (and other
>>time) in making the switch and testing it.
>>
>>Do the benefits of RMAN outweigh these costs (and the risks associated
>>with the disruption of something as fundamental as backups?) Almost
>>certainly, but some figures in terms of DBA time saved, s
>
>Apparently, you don't know RMAN, otherwise you wouldn't have made
>those comments.
>The company still NOT using Rman is not be envied, as Oracle clearly
>considers RMAN to be core technology, interweaving it more and more
>with the core database product.
>
is not the point. The point is that management need to be persuaded that
it is worth doing and shouting "this is the best technology" or "vendor
x says this is the best technology" is not the way to do it.
>As for your remarks on not upgrading: do you recommend companies to
>dump their support contract if they don't plan to upgrade.? Probably
>you don't. But if they don't upgrade, what is the value of their
>support contract?
>
I don't recommend anything. I was talking about the arguments that need
to be addressed to persuade people to upgrade.
>All in all I get the feeling you will soon get trivialized, as you
>don't seem to realize once you stop keeping your skills current,
>inevitably *you* will be dumped sometime, and stay unemployed, as your
>cv qualifies you as a dinosaur.
On the contrary, someone who can't read what is written, makes up
information which is absent, clearly has no understanding of how
business works, and jumps into personal insult without provocation is
much more likely to be "trivialized".
Specifically, you don't know what my attitude to keeping my skills up to
date is, and you haven't seen my CV so your ludicrous last paragraph is
worthless.
You seem to working very hard to outdo Daniel Morgan as an irritating,
knee-jerk bigot.
DA Morgan
>> Here are some recent examples ...
>>
>> - I had to pull teeth to get my current company to switch to RMAN.
>
> Why should they switch to RMAN? The currently backups presumably work
> fine, and there will be a significant cost in terms of DBA (and other
> time) in making the switch and testing it.
>
> Do the benefits of RMAN outweigh these costs (and the risks associated
> with the disruption of something as fundamental as backups?) Almost
> certainly, but some figures in terms of DBA time saved, shorter backup
> windows, faster restores etc, would help the argument.
You lack of experience and knowledge shine through the above like a
laser beam on a dark night. What cost or risk in implementing RMAN?
RMAN has been around for generations of Oracle and it is now the best,
safest, fastest, and least cost back up strategy available.
Oracle has put forth a no-brainer tool, no brainer in the same way that
implementing LMTs and UNDO are no-brainers. I've yet to see a single
argument put forward not to move to RMAN if it is available for the
version being discussed.
DA Morgan
>> Apparently, you don't know RMAN, otherwise you wouldn't have made
>> those comments.
>> The company still NOT using Rman is not be envied, as Oracle clearly
>> considers RMAN to be core technology, interweaving it more and more
>> with the core database product.
>>
> I do know RMAN and think all oracle sites should be using it, but that
> is not the point. The point is that management need to be persuaded that
> it is worth doing and shouting "this is the best technology" or "vendor
> x says this is the best technology" is not the way to do it.
As a DBA do you ask management to sanction using LMT rather than DMT? Do
you ask management if it is Ok to use UNDO rather than RBS? I don't like
your argument because as a DBA you are supposed to be a subject matter
expert. If you have tested RMAN and know how to implement it then you
implement it just as you would any other similar question within your
realm of expertise.
> Specifically, you don't know what my attitude to keeping my skills up to
> date is, and you haven't seen my CV so your ludicrous last paragraph is
> worthless.
>
> You seem to working very hard to outdo Daniel Morgan as an irritating,
> knee-jerk bigot.
Maybe both Sybrand and I, when reading English, come to the same
conclusion about the thoughts being communicated. I too read what you
wrote and assumed you to be paleolithic. You might wish to consider that
a communication involves more than what you intend. It involves what the
receiving party believes was communicated.
hpuxrac
> RMAN has been around for generations of Oracle and it is now the
best,
> safest, fastest, and least cost back up strategy available.
RMAN when it first came out was buggy and unuseable. People remember
things like that.
My opinion ... until 9i and the ability to handle control file
implications better it was still unsuited to many if not most oracle
installations.
If you already had a viable working backup solution, many people have
delayed implementing it.
Even now I doubt if RMAN is used at more than 40% of customer sites if
that.
Now I am using RMAN. But many people are not and do not use it
currently.
"Around for generations" ... what are you smoking?
premme...@hotmail.com
easily convince management? When it comes to backup/recovery, Data
Centers have their standards. Unless management knows Oracle Database
Amdinstration or let you make decision, it is hard battle. If people
are used to some tool for backup such as SQL Backtrack, they don't
want to switch. If data center does not kbow about ASM, it can take a
very long time to
convince. It is like if some data center standardized on using HP UNIX,
they do not like using anything else such as Solaris, Linux etc even if
those brands may be cheaper.
DBA's had hard job, on one hand they have to keep with latest and
greatest Oracle skills, on the other hand
depending on where they work, they may not be able to use most of new
features.
I have been wanting to use Oracle Application Server in some project,
but my company uses WebLogic.
Prem
DA Morgan
I remember many things from the past including the 60s. But time
moves on ... things change ... and I no longer hold the Oracle 7
installer against Oracle. Time to let go of the first release of
RMAN too. ;-)
DA Morgan
There is a big difference between buying a new product and paying for
licenses ... and using what you've already paid for. Who asks management
permission to create an index? Do you get specific permission to use
SQL*Loader? How about vi?
Noons
>
> As a DBA do you ask management to sanction using LMT rather than DMT?
Do
> you ask management if it is Ok to use UNDO rather than RBS? I don't
like
> your argument because as a DBA you are supposed to be a subject
matter
> expert. If you have tested RMAN and know how to implement it then you
> implement it just as you would any other similar question within your
> realm of expertise.
Very contingent on the type of site, Daniel.
I am currently working at one that has the dubious "advantage"
of being on DMT + RBO with 9ir2!!!!
The previous dba - for whatever reasons that don't matter here -
decided
to leave the schema and the data at V7 while upgrading the db
software to 9ir2. WTF for begs explaining, but let's ignore it...
With the result that I now have a headache of dinossauric proportions!
And an app that "only works with RBO", or so the development
manager has been led to believe.
So, my mission (should I accept it) is to:
1) convince the DM that contrary to what the previous dba told him,
CBO in 9ir2 is darn good.
2) convince the DM to do testing of the app with 9ir2 + LMT + CBO.
3) convince the CTO that all this is worth doing when all
he hears is that "MySql is free therefore must be good".
4) find a slot somewhere in a 24X7X365 system to do the necessary
manipulations to get rid of the old stuff and put in the new
with minimal disruption.
5) write a plan to implement 4) above in such a way that if ANYTHING
goes wrong, even the proverbial pointy-hair damager can set
things back the way they were.
Meanwhile, the backups are churning along with good old crontab/cpio.
Do you think for a second I'm going to touch them AT ALL before
I fix all the other crap?
NBW!!! (No-Bloody-Way!!!)
Life isn't always linear, my friend...
Jack
"Noons" <wizo...@yahoo.com.au> wrote in message
news:1113812644....@z14g2000cwz.googlegroups.com...
> DA Morgan wrote:
>>
>> As a DBA do you ask management to sanction using LMT rather than DMT?
> Do
>> you ask management if it is Ok to use UNDO rather than RBS? I don't
> like
>> your argument because as a DBA you are supposed to be a subject
> matter
>> expert. If you have tested RMAN and know how to implement it then you
>> implement it just as you would any other similar question within your
>> realm of expertise.
>
> Very contingent on the type of site, Daniel.
> I am currently working at one that has the dubious "advantage"
> of being on DMT + RBO with 9ir2!!!!
> The previous dba - for whatever reasons that don't matter here -
> decided
> to leave the schema and the data at V7 while upgrading the db
> software to 9ir2. WTF for begs explaining, but let's ignore it...
>
> With the result that I now have a headache of dinossauric proportions!
> And an app that "only works with RBO", or so the development
> manager has been led to believe.
>
> So, my mission (should I accept it) is to:
> 1) convince the DM that contrary to what the previous dba told him,
> CBO in 9ir2 is darn good.
As very known, there are still some issues f.ex with complex views.
Jack
Jack
From: "DA Morgan" <damo...@x.washington.edu>
Newsgroups: comp.databases.oracle.server
Sent: Monday, April 18, 2005 9:11 AM
Subject: Re: Are Oracle DBAs trivialized?
> There is a big difference between buying a new product and paying for
> licenses ... and using what you've already paid for. Who asks management
> permission to create an index? Do you get specific permission to use
> SQL*Loader? How about vi?
> --
> Daniel A. Morgan
> University of Washington
> damo...@x.washington.edu
> (replace 'x' with 'u' to respond)
I thinh RMAN would like to use tape-robot also which is operated an other
company.
So you surely need management aproval.
Or if you are using that near-on-line 4TB-system, you will need also
somekind aproval/agreements/contracts.
Daniel you are living in some kind of fantacy world.
Jack
Noons
>
>
> As very known, there are still some issues f.ex with complex views.
>
Good point. But those are already hinted to death in our
current code. And a few tests of some complex ones resulted
in CBO passing with flying results. I feel confident
we can take advantage of it and iron out the few
remaining "weird" ones.
--
Cheers
Nuno Souto
in sunny Sydney, Australia
wizo...@yahoo.com.au.nospam
Jack
Would you give some hints of those hints?
(Even every situation/issue is different.)
Jack
Noons
>
> Would you give some hints of those hints?
> (Even every situation/issue is different.)
>
Nothing fancy, remember: it's RBO.
Having said that, some of the plans are
scary! I've never seen before
some of the paths we get...
Send me a real mail to the addie below
and I'll send you an example.
DA Morgan
Your points are good and you have my sympathy. Don't they make test
boxes in your country? ;-)
Anyway that problem will go away if they ever decide to upgrade to 10g
or above. They'll look pretty funny demanding DMTs.
DA Morgan
> Jack apparently said,on my timestamp of 18/04/2005 9:21 PM:
>
>>
>> Would you give some hints of those hints?
>> (Even every situation/issue is different.)
>>
>
> Nothing fancy, remember: it's RBO.
> Having said that, some of the plans are
> scary! I've never seen before
> some of the paths we get...
> Send me a real mail to the addie below
> and I'll send you an example.
Me too please.
Thanks.
DA Morgan
No. But then I don't see any value in backing up to tape and have
said so here many times. Compare the cost of disk storage to tape
storage and disk wins. Compare the time required to back up to tape
versus disk and disk wins. So I am a loss to see why anyone would
back up to tape by choice.
But using your 4TB-system as an example. How long does it take you
to back up a 4TB system without RMAN? It can be done in minutes
(changed blocks) with the right tool. And what is that worth in time
and money?
hpuxrac
> licenses ... and using what you've already paid for. Who asks
management
> permission to create an index? Do you get specific permission to use
> SQL*Loader? How about vi?
Are you seriously asking if a DBA needs management buyin before testing
and implementing a significant change in the backup and recovery
strategy?
Do you not understand that there is a some small difference between
creating an index and switching backup and recovery strategies?
hpuxrac
> licenses ... and using what you've already paid for. Who asks
management
> permission to create an index? Do you get specific permission to use
> SQL*Loader? How about vi?
Are you seriously asking if a DBA needs management buyin before testing
Koki Sarmiento
DA Morgan wrote:
> No. But then I don't see any value in backing up to tape and have
> said so here many times. Compare the cost of disk storage to tape
> storage and disk wins. Compare the time required to back up to tape
> versus disk and disk wins. So I am a loss to see why anyone would
> back up to tape by choice.
Hi,
How exactly is disk storage cheaper than tape storage? I don't think
so. Could you explain why?
Just asking.
Regards
Koki
DA Morgan
Go to CDW and compare the cost per GB of tape versus IDE drives.
One can rotate disks in and out just as easily as tape and the
disks are both larger and faster.
Jim Smith
<damo...@x.washington.edu> writes
>Jim Smith wrote:
>
>>> Here are some recent examples ...
>>>
>>> - I had to pull teeth to get my current company to switch to RMAN.
>> Why should they switch to RMAN? The currently backups presumably
>>work fine, and there will be a significant cost in terms of DBA (and
>>other time) in making the switch and testing it.
>> Do the benefits of RMAN outweigh these costs (and the risks
>>associated with the disruption of something as fundamental as
>>backups?) Almost certainly, but some figures in terms of DBA time
>>saved, shorter backup windows, faster restores etc, would help the
>>argument.
>
>You lack of experience and knowledge shine through the above like a
>laser beam on a dark night.
You really are the most appalling ignoramus. Where is the lack of
experience or knowledge in knowing that actions which cost money and
introduce risk need to be analysed and justified in business terms as
well as technical terms.
>What cost or risk in implementing RMAN?
The cost of the time to implement and test it, and the risk of breaking
a currently working, *critical* component.
>RMAN has been around for generations of Oracle and it is now the best,
>safest, fastest, and least cost back up strategy available.
>
Prove it. Its probably true, but going to management and saying "I am
the DBA and rman is the solution to all our problems" is not the way to
go about it.
>Oracle has put forth a no-brainer tool, no brainer in the same way that
>implementing LMTs and UNDO are no-brainers. I've yet to see a single
>argument put forward not to move to RMAN if it is available for the
>version being discussed.
--
Jim Smith
<damo...@x.washington.edu> writes
>Jim Smith wrote:
>
>>> Apparently, you don't know RMAN, otherwise you wouldn't have made
>>> those comments.
>>> The company still NOT using Rman is not be envied, as Oracle clearly
>>> considers RMAN to be core technology, interweaving it more and more
>>> with the core database product.
>>>
>> I do know RMAN and think all oracle sites should be using it, but
>>that is not the point. The point is that management need to be
>>persuaded that it is worth doing and shouting "this is the best
>>technology" or "vendor x says this is the best technology" is not the
>>way to do it.
>
>As a DBA do you ask management to sanction using LMT rather than DMT? Do
>you ask management if it is Ok to use UNDO rather than RBS? I don't like
>your argument because as a DBA you are supposed to be a subject matter
>expert. If you have tested RMAN and know how to implement it then you
>implement it just as you would any other similar question within your
>realm of expertise.
>
Any activity which costs money and introduces risk above a certain level
needs management approval in every business I've ever worked in. Perhaps
you have unlimited budgets?
>
>Maybe both Sybrand and I, when reading English, come to the same
>conclusion about the thoughts being communicated. I too read what you
>wrote and assumed you to be paleolithic. You might wish to consider that
>a communication involves more than what you intend. It involves what the
>receiving party believes was communicated.
Reading before you post has never been your strong point.
Sybrand Bakker
<j...@jimsmith.demon.co.uk> wrote:
>You really are the most appalling ignoramus. Where is the lack of
>experience or knowledge in knowing that actions which cost money and
>introduce risk need to be analysed and justified in business terms as
>well as technical terms.
I don't know why you keep repeating this mantra.
The only reason I can imagine is that you really don't know RMAN,
otherwise you wouldn't have made those remarks
>
>>What cost or risk in implementing RMAN?
>
>The cost of the time to implement and test it, and the risk of breaking
>a currently working, *critical* component.
In my experience likely *every hand made* backup script is less robust
than RMAN. So when you are talking about *critical* components,
chances are you might already have one. You aren't aware of that
because self-developed solutions are often justed *trusted* instead of
*tested*, and it often is quite easy to break them. One little example
is a script where the list of files to be backuped had to be
maintained in a separate configuration file. As you claim you know
everything so much better, you could probably, as a quiz question,
identify why this is a crap solution, and identify the costs involved.
DA Morgan
> Any activity which costs money and introduces risk above a certain level
> needs management approval in every business I've ever worked in. Perhaps
> you have unlimited budgets?
I agree so lets examine your statements.
RMAN doesn't cost money. If you have it you have already paid for it.
And one should probably point out that when management approves the
purchase of a software package, say Oracle 10g, they have approved
purchasing the entire package ... they didn't say well we'll take it
all but not interMedia.
What risk? I think you introduce a far higher degree of risk by not
using RMAN and I think there is a huge amount of documentation that
supports this that your Oracle SE would be happy to provide to you.
There is a risk in doing something ... that I will grant. But there
is also a risk in doing nothing. To (excuse me for the word) pretend
there is no risk involved in doing nothing is methinks a mistake.
At least that is my opinion and I'm sticking with it. ;-)
Joel Garry
>Any activity which costs money and introduces risk
>above a certain level needs management approval
>in every business I've ever worked in. Perhaps
>you have unlimited budgets?
The mistake implicit here is that management has costed deferred
maintenance. I haven't seen a case yet where they've fallen way behind
in technology and accounted for that. The usual case is they costed
the project when it was new, depreciated as it went through its life
cycle, and thought they were reducing costs by not keeping it up to
date.
Really, a failure of strategic planning more than cost accounting,
since the costs were kept down. But incorrect costing in real terms
(as opposed to accounting fiction), since management didn't get a
report that says "these are the deferred costs that are adding up." So
Jim, all those businesses you've been in, have you seen such a report?
Should you have, under your own "needs management approval" definition?
It's tough for management to delegate to DBA's these sorts of
decisions, especially since "business decisions" are too often framed
in terms of accounting, and it is difficult to include non-accounting
concepts like keeping workers happy and evaluating how critical
technology is to a business. So all we can do is throw around
buzzwords like... well, you know, the stuff they read in CIO magazine
or whatall, and hope that can be called business analysis and that we
can slip in the right stuff when nobody is looking.
Certainly, situations like Noons describes can in no way be considered
good business planning. Management doesn't even know it is bad (well,
hopefully Noons is dealing with that). As always, we can all blame
the previous DBA! :-)
jg
--
@home.com is bogus. "A lot in there to chew on"? Sounds like dead
meat to me:
http://news.com.com/Microsoft+to+license+test+software+for+real-world+use/2100-1012_3-5675172.html?tag=nefd.top
.
Jim Smith
<pos...@sybrandb.demon.nl> writes
>On Mon, 18 Apr 2005 20:47:37 +0100, Jim Smith
><j...@jimsmith.demon.co.uk> wrote:
>
>>You really are the most appalling ignoramus. Where is the lack of
>>experience or knowledge in knowing that actions which cost money and
>>introduce risk need to be analysed and justified in business terms as
>>well as technical terms.
>
>I don't know why you keep repeating this mantra.
Because its better than the mantra, "RMAN is better because I say so."
>The only reason I can imagine is that you really don't know RMAN,
>otherwise you wouldn't have made those remarks
This is so stupid. Can you explain your reasoning? How does basic common
sense imply I don't know RMAN.
>>
>>>What cost or risk in implementing RMAN?
>>
>>The cost of the time to implement and test it, and the risk of breaking
>>a currently working, *critical* component.
>
>In my experience likely *every hand made* backup script is less robust
>than RMAN.
Not everyone who isn't using RMAN is using hand-made scripts. There are
other solutions, but in any case a carefully written script by an
experienced DBA can certainly be robust. DBAs were writing backup
scripts for a long time before RMAN was introduced and probably got
quite good at it.
>So when you are talking about *critical* components,
>chances are you might already have one.
You misunderstand the meaning of "critical". It means that a component
is (very) important, not that it is more likely to fail.
> You aren't aware of that
>because self-developed solutions are often justed *trusted* instead of
>*tested*,
Someone who hasn't tested their current solution is also unlikely to
test their rman solution.
>and it often is quite easy to break them.
>One little example
>is a script where the list of files to be backuped had to be
>maintained in a separate configuration file. As you claim you know
>everything so much better, you could probably, as a quiz question,
>identify why this is a crap solution, and identify the costs involved.
>
It is probably equally easy to develop a solution using rman which is
fatally flawed.
>
>
>
>--
>Sybrand Bakker, Senior Oracle DBA
--
Jim Smith
<damo...@x.washington.edu> writes
>Jim Smith wrote:
>
>> Any activity which costs money and introduces risk above a certain
>>level needs management approval in every business I've ever worked
>>in. Perhaps you have unlimited budgets?
>
>I agree so lets examine your statements.
>
>RMAN doesn't cost money. If you have it you have already paid for it.
>And one should probably point out that when management approves the
>purchase of a software package, say Oracle 10g, they have approved
>purchasing the entire package ... they didn't say well we'll take it
>all but not interMedia.
does. You may also have to upgrade backup software for RMAN
compatibility (on the assumption that if you are not keeping up to date
generally, your backup software is also out of date), buy more disks,
more tapes, whatever.
>
>What risk? I think you introduce a far higher degree of risk by not
>using RMAN and I think there is a huge amount of documentation that
>supports this that your Oracle SE would be happy to provide to you.
>
upgrades, platform migrations, DBMS migrations - but the ones that are
most likely to cause sleepless nights are changes to backups. The cost
of failure is huge, and there is always the fear that there is something
wrong that you won't discover until you have to restore... I have yet to
come across a site which has a really satisfactory test backup
environment.
>There is a risk in doing something ... that I will grant. But there
>is also a risk in doing nothing. To (excuse me for the word) pretend
>there is no risk involved in doing nothing is methinks a mistake.
>
Agreed, there should always be a cost of doing nothing section.
Unfortunately "Well, duh", isn't very persuasive.
>At least that is my opinion and I'm sticking with it. ;-)
--
Noons
>
> Your points are good and you have my sympathy. Don't they make test
> boxes in your country? ;-)
>
I wish I had $1 for every time I've found a
test box that is *not quite* the same as production...
;)
> Anyway that problem will go away if they ever decide to upgrade to
10g
> or above. They'll look pretty funny demanding DMTs.
It's going away NOW, if I get my way...
D Rolfe
>
>
> I wish I had $1 for every time I've found a
> test box that is *not quite* the same as production...
> ;)
>
>
You think that's bad? I once worked at a place where the backup/failover
system was VMS and our production system was Solaris....
It gets better. Our hardware guy had a 'special relationship' with
Digital, which meant we must have been the only site in America that was
attempting to migrate from Unix to VMS.
Our office was located at the Northern end of the Calvaras earthquake
fault in California, which means any disaster plan has to assume nuclear
levels of destruction. Our backup system was located 50 miles away.
Actually 'located' is the wrong word to use. There was an empty computer
room 50 miles away and DEC had 'promised' our manager they'd give him
the first box off the production line if he ever needed it. As plan 'B's
go it's up there with the safety systems on the Titanic...
David Rolfe
Orinda Software
Dublin, Ireland
DA Morgan
>> In my experience likely *every hand made* backup script is less robust
>> than RMAN.
>
> Not everyone who isn't using RMAN is using hand-made scripts.
But those that aren't are paying a lot of money to avoid using that
which their company already purchased.
> There are
> other solutions, but in any case a carefully written script by an
> experienced DBA can certainly be robust. DBAs were writing backup
> scripts for a long time before RMAN was introduced and probably got
> quite good at it.
'Probably' is an excellent choice of words. I have yet to see the
DBA whose script was put into change management and given to an
independent team for testing and certification. And those that do
work ... work only so long as the DBA has tribal knowledge of what
is where and manually changes it each time a change is made to the
database.
Want to know how many times I have seen these carefully crafted
scripts fail when a datafile was moved or added?
>> So when you are talking about *critical* components,
>> chances are you might already have one.
>
> You misunderstand the meaning of "critical". It means that a component
> is (very) important, not that it is more likely to fail.
Your definition but not one generally in use. How about critical as in
a show-stopper if it doesn't work perfectly. It may or may not be likely
to fail. But when it does, such as an inability to restore, the system
stays down.
>> You aren't aware of that
>> because self-developed solutions are often justed *trusted* instead of
>> *tested*,
>
> Someone who hasn't tested their current solution is also unlikely to
> test their rman solution.
Agreed. Those that don't test deserve their own special place in hades.
>> and it often is quite easy to break them.
>
>> One little example
>> is a script where the list of files to be backuped had to be
>> maintained in a separate configuration file. As you claim you know
>> everything so much better, you could probably, as a quiz question,
>> identify why this is a crap solution, and identify the costs involved.
>>
> It is probably equally easy to develop a solution using rman which is
> fatally flawed.
But there are things I can do in RMAN you can not do with a script. And
lets start by discussing block recovery.