Question How much free information is too much?

[TomBascom]

Nice.

I ran it without any issues.

I was, of course, dismayed to crack it open and see that the code is full of &ANALYZE-SUSPEND hooey. But aside from that I like it

 

[TheMadDBA]

lol... I actually have learned to like using the AppBuilder for development because I can jump back and forth between functions and procedures so quickly. Still not a fan of trying to create actual GUI applications though.

I bet you loved my various naming conventions too

Thanks

 

[TomBascom]

I went blind as soon as I saw the &ANALYZE-SUSPEND stuff.

I do like the output. It is crisp and to the point.

 

[joey.jeremiah]

i definitely think that - everything that can be done automatically should be!

cool idea!

 

[Cringer]

New version works a treat, and so much for being well tuned
HIGH Database 300 Tables with storage issues Some of your tables are either in Type I areas and/or do not have proper record per block settings.
HIGH Database 174 Indexes flagged for rebuild Some of your indexes need to be rebuilt or compacted to improve performance.

 

[Cringer]

A suggestion, and I don't know if it's possible, but are you able to identify LOBs and the area they are assigned to? I've not looked to be honest. Our LOBs are in data areas and urgently need moving to separate areas, so I know it's there in ours so might be a good one to base any code to handle it on.

 

[TheMadDBA]

I will add the LOBs to the to-do list. I know we have less information about them than tables but storage areas are certainly there.. basically warning if they are in the same areas as other non LOBs. To be honest I think most of the issues are going to be with maintenance and not so much on the performance side.

You can run this for now (stolen from a KB) to find out where they are for sure.

FOR EACH _storageobject WHERE _storageobject._object-type = 3 NO-LOCK:
FIND _file WHERE _file-num = _storageobject._object-associate NO-LOCK.
FIND _field OF _file WHERE _fld-stlen = _storageobject._object-number NO-LOCK.

DISPLAY _file._file-name
_field._field-name
_storageobject._object-number
_storageobject._area-number.
END.

I will take another look at your dbanalys file and double check the suggestions. They should be sorted in order of relative importance (most active at the top). The ones that have little or no activity won't really matter unless you have different times of the month/year that they suddenly become very active.

Thanks again for testing this out for me

 

[Cringer]

No problem at all - my output is attached.

 

[TheMadDBA]

New version that should trim down your suggested list of table moves a bit. You have quite a few tables with low record counts (or zero). Tweaked the logic a bit to try and filter those out.

 

[Cringer]

Thanks. Will give it a go.
A full D&L is on the cards for the not too distant future and this is really going to help me.

 

[c_verbiest]

I added it as an optional component to https://github.com/cverbiest/dbanalys-charts . I hope this is OK, if not I'll remove it again.

 

[tamhas]

Both might consider OE Hive for more visibility.

 

[TheMadDBA]

That will be fine for now. I am still working on a few things and plan to officially publish it soon. I do think it has value "as is" though.

I plan on putting the code and documentation on OE Hive, GitHub, PSC Communities and probably a few others.

At some point I think it would be a good idea to integrate the two programs a little more tightly so that one run can generate the diagnostic information and the charts.. probably with some HTML linking between the two.

 

[TheMadDBA]

Bumping this again... I have my site up and running now. It has a few more utilities available in addition to the Advisor (generic table caching, fragmentation map for Type I tables and some tools that integrate the Profiler with the _UserTableStat and _UserIndexStat VSTs).

Still have more pages and information to add and a few more spell checks and grammar checks too.

When you get a chance let me know what you think... It should be relatively browser/tablet/phone friendly.

http://www.themaddba.com

Thanks

 

[Cringer]

Funny this should appear - just set up a DBANALYS schedule in production meaning I have a weekly report, making this sort of info more easy to test

 

[Cringer]

DB Uptime is too short at the moment, but will run again next week. One thing it did remind me is that I forgot to proaiw the database when I restarted it. Reminding me once again that it's ridiculous that you can't set that to autostart in OEM.
I notice you've taken out the last 2 parameters. I think that's probably a good thing. What did you do to make the change?

 

[Cringer]

So reading your comments on the AIW, I see "you can check the appropriate box in OpenEdge Explorer/Management". Is this this? After image process:

 

[Rob Fitzpatrick]

When you get a chance let me know what you think... It should be relatively browser/tablet/phone friendly.

Great job on the site. Lots of good information and well presented.

 

[Cringer]

Yup totally agree. I've linked to it from my website now as well.

 

[TheMadDBA]

So reading your comments on the AIW, I see "you can check the appropriate box in OpenEdge Explorer/Management". Is this this? After image process:

That would be the one. The exact description seems to vary depending on the version. Some places actually include (AIW) after the description. I think that should be the standard.

Thanks for the input guys. I think I still have a long way to go to get the site to the level I want it to be, but it is a start I guess.

As far as removing the options for the table scans.... Most of the logic in there really only applies to Type I areas. I split that out and merged it with my Block Mapper tool.