COLLECT bug?

Anybody else encountered a problem using COLLECT in a query? Seems to work for me in Sql+, but not SqlDev 1.5.0.52/1.5.3
select tablespace_name, collect(table_name) as table_name_coll
from all_tables
group by tablespace_name;
Error starting at line 1 in command:
select tablespace_name, collect(table_name) as table_name_coll
from all_tables
group by tablespace_name
Error report:
SQL Error: ORA-00932: inconsistent datatypes: expected NUMBER got -
00932. 00000 - "inconsistent datatypes: expected %s got %s"
*Cause:
*Action:

We have closed the bug mentioned as "Not a Bug". The CAST function must be used in conjunction with the the COLLECT function.
From the Database 11g documentation:
"COLLECT is an aggregate function that takes as its argument a column of any type and creates a nested table of the input type out of the rows selected.
To get accurate results from this function you must use it within a CAST function."
Sue

Similar Messages

Smart Collection bug with Title doesn't contain rule

Hi,
There are various articles on smart collections that check if the "Title" field is filled or not. Because the "Title" field doesn't have an "is empty" like function, the sugested solution is to use the rule:
Title   doesn't contain   a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z
But there is a bug in this rule. I found that the letter 't' is causing a problem. The letter 't' seems to include all characters.
Situation:
Four images that have a their Title field as follows:
Image 1:
Image 2: 2014-01 - Street Photography.dng
Image 3: 2014-01
Image 4: Some Image
(Image 1 has an empty title)
Four smart collections with the following rules:
Collection 1: Title   doesn't contain   a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z
Collection 2: Title   doesn't contain   0,1,2,3,4,5,6,7,8,9
Collection 3: Title   doesn't contain   a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,u,v,w,x,y,z
Collection 4: Title   doesn't contain   t
My expected result of the smart collections is:
Collection 1: Image 1 and Image 3
Collection 2: Image 1 and Image 4
Collection 3: Image 1 and Image 3
Collection 4: Image 1, Image 3 and Image 4
The actual result of the smart collections is:
Collection 1: Image 1
Collection 2: Image 1 and Image 4
Collection 3: Image 1 and Image 3
Collection 4: Image 1
I'm using Lightroom 5.7 on a MacBook Pro 17" (mid 2010), i7 @ 2.66GHz, 8GB RAM and 2 x 240GB SSD's running on OS X 10.10.1
My questions are:
Can anyone, specially from the Adobe Lightroom team, confirm this?
How to solve this issue?
With kind regards,
Dennis
www.idemi.nl

That is a bizarre bug. There is an even simpler way to reproduce the problem: Have two pics, one with title "a" and the other with title "t". The criterion "Title Contains a" correctly shows just the first picture, but the criterion "Title Contains t" shows both.
Searching with smart collections and filters has numerous bugs, but this one is by far the most surprising. As a geek programmer, I'd love to know what causes this!
Adobe rarely reads this forum -- I suggest you post this in the official Adobe feedback forum, which Adobe does faithfully read (but only infrequently responds).

Smart Collections bug

A smart collection of:
rating is not * (1 star)
should return images including those with no stars, but does not.
(It only returns items with 2,3,4 or 5 stars).
Mike.

Known bug.
http://forums.adobe.com/thread/657891

LR3 Smart Collection Bug

Hey everyone,
I'm having issues with my smart collections in LR3.
When I set a rating to "is not" it only includes ratings higher than that. It doesn't include photos, not rated or rated lower.
Anyone else with this issue?
Cheers

Change the logic:
And report the bug.

Ten Smart Collection bugs/changes

After upgrading the catalog from v2 to v3 I have compared the smart collection counters for every SC and found a lot of differences. So, you may need to check your own SCs or rewrite some rules. Here's a list of bugs and changes in logic I was able to find.
"Rating is not X" does not include unrated photos
Be it a bug or change in logic, but LR2 would include unrated photos in this case. This has also been discussed here: http://forums.adobe.com/message/2876677 .
"Rating is 0" includes all photos
This is definitely a bug, as now you cannot catch your unrated images without tricks (e.g. "None: rating >= 1").
+ modifiers reversed role
In LR2 "+abc" meant "starts with abc" and "abc+" meant "ends with abc". In LR3 these have reversed. More than that, in LR2 one could use "+abc+" to identify the word "abc"; in LR3 this doesn't work anymore, since "contains +abc+" is equivalent to "contains abc".
! modifier doesn't work
the ! (not) modifier just doesn't work here. The smart collection just hangs. This also causes other SCs to not update their counts after start-up. Try "keyword contains !abc".
Dash ("-") is now treated as a word delimiter
In LR2 "abc-xyz" was treated as a complete word. LR3 treats it as two words: "abc" and "xyz". Not a bug, but some SCs will have to be rewritten.
"Caption is empty" doesn't work for some images
LR3 catalog has a new table in its schema: AgLibraryIPTC. This table appears to be a sort of a cache for captions and copyrights (not to confuse with copyright status). This particular SC rule seems to search in that table in instead of full xmp. However, if an image didn't contain neither a caption nor a copyright in LR2 catalog, after conversion it won't be added to the AgLibraryIPTC table. This, in its turn, will cause the "Caption is empty" rule to fail even though a photo does indeed NOT have a caption.
"Title contains t" includes all image from library
Now this is a really strange one, as I don't have any titles on any of my photos. I tried every other letter and digit, but only "t" fails the check. "Title doesn't contain t" also fails, not including any photo at all, although none of the photos have any title. Again, "t" is the only wrong letter. "T" is often used as an alias in sql joins... could it be the problem?
"Title contains <space>" hangs
Just like the "!" modifier above, this rule hangs and blocks other "healthy" SCs to not recount themselves after start-up. You need to type in a single space character into the criteria filed.
"Label Text" is case sensitive
It has been so in LR2 too, but I think a user should be freed from the need to think about case-sensitivity.
"All searchable metadata" doesn't search in copyright.
Note that "All searchable Text" DOES search there, but isn't the copyright field metadata? Note that LR didn't search in copyright with both rules, so this is a kind of improvement, after all.

Just for your information:
"Rating is not X" does not include unrated photos Fixed in 3.2 RC.
"Rating is 0" includes all photos somewhat fixed, but with issues
+ modifiers reversed role fixed, but with issues
! modifier doesn't work Fixed in 3.2 RC.
Dash ("-") is now treated as a word delimiter Fixed in 3.2 RC.
"Caption is empty" doesn't work for some images Fixed in 3.2 RC.
"Title contains t" includes all image from library Fixed in 3.2 RC.
"Title contains <space>" hangs not fixed
"All searchable metadata" doesn't search in copyright. not fixed, but might not be a bug
"Label Text" is case sensitive not fixed, but might not be a bug

Keywording/Collection Bug?

I had a set of 12 images in a collection. I selected all twelve and applied a keyword to them (eg. "AAAAA"). Then I hit command-L to switch off the catalog filters. I then went to the Keyword list panel and confirmed that the number of images with the keyword was indeed 12. I then clicked on the right arrow for the keyword "AAAAA"
This should result in the filter bar popping up and I would expect to see my twelve images. Instead I only saw six and the Library Filter keyword for "AAAAA" has the number six by it. If I go to each individual file in its folder and check the keyword tags I correctly see the keyword "AAAAA" for each of them. The Catalog has "All Photographs" selected, so it is not as if I am only searching a subset of my catalog. If I go to the folder(s) where the files are located and individually remove and then re-add the keyword, I see the image I just "keyword-fixed" pop up in the filter search. So, it seems to me there is a bug when applying keywords to images in a collection.
Anybody else seen this? Some screen shots are on my blog.
Dave
www.davehartleyphoto.com

That is a bizarre bug. There is an even simpler way to reproduce the problem: Have two pics, one with title "a" and the other with title "t". The criterion "Title Contains a" correctly shows just the first picture, but the criterion "Title Contains t" shows both.
Searching with smart collections and filters has numerous bugs, but this one is by far the most surprising. As a geek programmer, I'd love to know what causes this!
Adobe rarely reads this forum -- I suggest you post this in the official Adobe feedback forum, which Adobe does faithfully read (but only infrequently responds).

Resource string for panel collection bug?

Hi all,
Resource string af_panelCollection.LABEL_STATUSBARITEM_SELECTED_ALL for panel collection doesn't work.
I believe this one is misspelled because all other resource strings that I use work fine.
Does anyone know real resource string for this?

Hi,
this string seems to be missing in PanelCollectionRenderer.class. I'll file a bug
Frank

Another collections bug?

Following on from this thread relating to incorrect debug session reporting of collections:
Re: Strange collection errors occuring on only some workstations (apex 2.2.
I am encountering more strange behaviour using collections. I hope its my own programming, however I am increasingly led to believe that something is not quite right within the APEX collections implementation.
I am working within a corporate environment, and my application works fine on the local desktop. However when I try and launch the application from within a Citrix Metaframe desktop, the following error INTERMITTENTLY occurs:
ORA-20104: create_collection_from_query_b Error:ORA-20101: Application collection exists
The url used to launch the application is the base url, so am I correct in thinking a new session should be created? This point is moot however, because on load of my application I explicitly delete all collections and the problem still occurs.
Perhaps the problem lies with using the Citrix desktop, however I think it is more likely that the differences in IE versions are part of this problem.
IE (desktop) version: 6.0.2800.1106.xpsp2.050301-1526CO
IE (citrix desktop) version: 6.0.2800.1106CO
So the obvious difference here is that I have XP-SP2 and the Citrix desktop is running 2000 SP4.
The other intermittent error I get is as follows:
ORA-20104: create_collection_from_query_b Error:ORA-00001: unique constraint (FLOWS_020200.WWV_FLOW_COLLECTIONS_UK) violated
The collection exists of one column and it is a GROUP BY query so duplicates are not possible. Is some other internal field acting as the primary key inside a collection? Or is the collection trying to append to the previous collection? Again, I am very confused because all collections are deleted 'on submit', before the new collection is created.
How can I view this 'FLOWS_020200.WWV_FLOW_COLLECTIONS_UK' value?
Can anyone point me in the right direction?

Update:
I have managed to pinpoint the exact sequence of event that causes this error:
ORA-20104: create_collection_from_query_b Error:ORA-00001: unique constraint (FLOWS_020200.WWV_FLOW_COLLECTIONS_UK) violated
Step taken to produce the error:
1.) Launch application from URL
2.) Submit Page 1 (returns a report on the same page)
3.) ORA-20104 as above.
Repeat ad nauseum.
I would like to reiterate that the APEX_COLLECTION.DELETE_ALL_COLLECTIONS_SESSION command is being run on load (even though it should not have to be run.) I know this is executing on load from the process success message.
Does the DELETE_ALL_COLLECTIONS_SESSION statement delete the actual collection or merely the reference to the collection? I must admit I am quite clueless when it comes to the internal workings of oracle.
My next step is to come in tomorrow morning and see if it fails the first time I run it.
If it does run correctly, I would assume that all collections have expired and that DELETE_ALL_COLLECTIONS_SESSION is not doing what it says it is.

Exported catalogue has lost some collections - bug?

Hi,
After recently exporting a catalogue (of all photos in that catalogue) and then importing the catologue into a fresh lightroom catalogue it has lost several (about a quarter) of the collections. This is a bit of a bugger as a fair bit of work went into creating them. Anyone got any clues as to how to prevent this?
Thanks,
Tom

Not so with web gallery collections or print collections. Export a catalog, then open it and they are missing.
See previous post:
http://forums.adobe.com/thread/692631?tstart=90
Makes syncing catalogs between computers difficult. I guess you can change the web gallery and print collections to regular collections, but why have them if they do not work as expected.
Thanks,
JB

Possible bug with Pick flag and Collections

Odd behaviours in v3.3 (Windows) with pick flags in Collections:
Create (or use) a Collection Set which contains a number of different collections
Assign Pick plag to one photo in each collection
Bug 1: Highlight Collection Set and note that none of the images shows the Pick Flag in the thumbnail view. Also Filtering this view with Pick flag fails to pick up the Flagged images
Now return to one of the Collections within the Set and create a Virtual Copy of one of the images that has a Pick flag. (Note virtual copy does not have a flag - which seems logical). Return to Collection Set and (Bug 2) note that Pick flag has disappeared from original image and now appears on Virtual copy.
Something wrong here somewhere I suspect.

lightshop wrote:
Odd behaviours in v3.3 (Windows) with pick flags in Collections:
Create (or use) a Collection Set which contains a number of different collections
Assign Pick plag to one photo in each collection
Bug 1: Highlight Collection Set and note that none of the images shows the Pick Flag in the thumbnail view. Also Filtering this view with Pick flag fails to pick up the Flagged images
Now return to one of the Collections within the Set and create a Virtual Copy of one of the images that has a Pick flag. (Note virtual copy does not have a flag - which seems logical). Return to Collection Set and (Bug 2) note that Pick flag has disappeared from original image and now appears on Virtual copy.
Something wrong here somewhere I suspect.
I could partially repeat bug 2.
I have Collection Set (Test) which contain to collections: "Collection 1" and "Collection 2". Both Collections contain 4 images.
I selected "Collection 1" and applied Pick flag to all those images. When I select Test (Collection Set) I got 8 images shown, which none had flags. Ok, this works as expected. I selected "Collection 1" and then selected 1 image and made Virtual copy from it. Lr show me now 4 images with Flag and 1 virtual copy without. Again, works as expected. I now select Test (Collection Set) and I get 9 images (as expected), but one of them, the virtual copy, has pick flag applied. Not as expected. If I go back to "Collection 1" I get four images with Pick flags and one virtual copy without. (as expected.)

Error in collect function

Hi,
I am using below query for collect
select account_id_N,collect(connection_id_v) from My_test_table group by account_id_n;
but its giving me error as below.
ORA-00932: inconsistent datatypes: expected NUMBER got -
00932. 00000 - "inconsistent datatypes: expected %s got %s"
*Cause:
*Action:
can anyone please help me to find the mistake.
Thanks.

The document clearly states you must use CAST with COLLECT.
http://docs.oracle.com/cd/B19306_01/server.102/b14200/functions024.htm#SQLRF06304
something like this.
create type emp_name as table of varchar2(100)
select deptno, cast(collect(ename) as emp_name) enam_list from emp group by deptno
/ Note: Collect function have some behavioral difference when used from SQL Plus and SQL Developer
COLLECT bug?

Speedgrade bug reporting...

Just a quick query for the Speedgrade devs- do you guys collect bug reports from the Adobe improvement program reports? I'm seeing a lot of crashes when editing Cineform footage, but only when Speedgrade exits. Speedgrade itself can be quite crashy, and sadly I lack the time to try and report them all. If you guys are grabbing data from the improvement program then that'll save a lot of hassle.

Hi,
Adobe SpeedGrade is participating at the improvement program. Find a list of all participating products here: http://www.adobe.com/special/apip/improvement.html
If you encounter a crash, I would recommend to also post a report to the forums. Probably a faster way to find a workaround/solution for the problem.
What system are you using (OS, graphics etc.) ?
SpeedGrade version ?
I would like to take a look at crashlogs and SpeedGrade logs. Sent you a PM with further instructions.
Dennis

Patch for modbin6-1.00.38.exe is out.. "2MBIOS.BIN can't open!" Solved.

Hi mates,
ive decide to make a patch for modbin6 1.00.38 to improve stability and minise risk of modding the BIOS. for users which is reading this post if never heard about MODBIN better avoid reading below. its nice program for basic BIOS mod, simple to use, but very danger in the same time couse mess with BIOS and inproperty using may couse disaster. im not gona advert or to invite persons to use it. just wana make this program to be used normally to bring stability and minise risk of making corrupted BIOSes.. for persons which is already using it and have experience with him or with person Enthu which is wana try it and they dont worry if mobo may caouse for trash. there is 2 reasons to making this patch.
1. couse sometimes im using it there is features on it wchich gives basic stuffs, unlocking hidden features locked by manifacturer(sometimes really usefull especially for DIMM Vcore..). so there is a nasty bug on it which gona explain later. most of persons who is using MODBIN know about it. so that patch will spend me alot of time and minimise risk of mistake during avoiding bug couse operation need several procedure and automatised will risk free and much fastern.
2. to bring ability for most stable operations to persons which is already using it when need.
but lets talk about the patch. everyone know that nasty issue with MODBIN(made it by Borg N1,part from his collection). bug is program wont safe BIOS file and quiting with Abnormal program termination.... gives this one message: !!!File
2MBIOS.BIN can't open!Thanks for using ModBin6 1.00.38 how nice ...
here is Borg N1 advices:
Quote from: Borg Number One on 16-July-05, 05:24:51
Hi.
It is a kind of sad fact that many users just know how to use the mouse and keyboard.
Well, you have to configure your config.sys (Win9x/ME) or config.nt to run MODBIN6
(e.g. MODBIN6 2.01.01 (06/25))
in Windows prompt/console without problems.
The " FILES " & " BUFFERS " values are the important values.
http://www.xtremesystems.org/forums/showthread.php?t=45538
http://www.xtremesystems.org/forums/showthread.php?t=63210&highlight=config.nt
+++ BNOBTC & BNOBLG +++
http://bnobtc.pix-art.com
Borg advices is good but unfortuanly doesnt fix that bug... seems author dont have much idea why this is happend...error is constant ... Pure DOS also can't help... i wana add few more tips to the Borg's ones to improve stability when using modbin, after that will talking about patch.
TIP: navigate to this folder and open config.nt file: %SystemRoot%\system32\config.nt
at the end of file add this ones:
dos=high, umb
device=%SystemRoot%\system32\himem.sys /CPUCLOCK:ON /HMAMIN=64
FILES=40
BUFFERS=40
STACKS=9,256 -----> Preserving 039A-045Eh 3,152 Sys 039Ah Stacks, 0552h DOSX
Also MODBIN doesnt work well with long filenames.... etc "document and settings\my profile\desktp\modblabla" if not gona work property and there is a big chance of failure will load but after saving will have problems to update BIOS "core"(original.tmp) so always use 8.3 syntax standart .... best way from root directory make subdir whcih name to no NOT exceed 8 symbols... etc C:\123,C:\MYFLASH,C:\ROFL
without TIPs it is really big gamble with BIOS....
Aslo that will prevent Stack Overflow error when BIOS is loading....
NOTE: To activate changes just quit from any DOS interpretators like(4nt,command.com,cmd and so on). and re-run it. Reboot OS is not requred.
here how to do:
Also when executing MODBIN put below executable file this ones without quotes "loadfix forcedos modbin6-1.00.38.exe"
and ofcourse nasty error is waiting us... "2MBIOS.BIN can't open!Thanks for using ModBin6 1.00.38"
Patch which iv maked gona RIP this error forever.
Download: modbin6_1.00.38
Download: MODBINPATCH
What patch need? and how to use it? I named patch "Dynamic Patch for modbin" Couse isnt make modification on MODBIN program but momory his parent process when is loaded into memory.
tips suggested below is recommended but not nesecery for the patch.. patch need MODBIN and your BIOS file (modbin6-1.00.38.exe and mybios.bin) to be copied to systemroot drive(usually C) in folder MSI etc put modbin6-1.00.38.exe and mybios.bin(BIOS which you gona mod) to C:\MSI
Patch make checking if folder doesnt match will not continue....
How to use it..... Copy modbin6-1.00.38.exe and mybios.bin to C:\MSI, then from command prompt navigate to C:\MSI and execute "loadfix forcedos modbin6-1.00.38.exe" after that load your BIOS file... when loading is complete navigate to your patch folder and execute patchMSI.exe.... takes 10 seconds to complete after that make active "white" window and enter "Q" then hit ENTER key both windows will be closed, after that return to your modbin window, make what you gona make and save BIOS file like normal procedure... Method is founded and tested by me from years below never let me down(never brings corrupted BIOS), but just today ive made patch couse was too lazy... also Borg's MODBIN can work with most of all chipsets which is using AWARD Modular based... NOTE: K9 isnt tested. the most of here you can view how exacly to do it and differents with and without patch:
here is one more think some time maybe usefull from me:D
Here is 1st little short trick:(no analog anywhere, like patch only here can be found if accepted):
How to reset BIOS if you have nothink? Examples if you are administrator/PC-support or any and must fix same PC but the BIOS has an password?! And you dont have permision to open Case to reset CMOS(couse PC has an sticker with Warantly (and removing that sticker couse lost pc warrantly) You need access to BIOS to re-check,configure somethink there but its password protected. anyway there could be many situation. what we need? nothink special just bootable DOS(or analog- PC-DOS or whatever,doesnt matter where will boot(cd,flash disk,hdd)), no extended commands.[io.sys(jo.sys),command.com] - files requred to boot Here we go:
boot from desired device.when you see command prompt etc.: "A:\,C:\"
We will write a program without any languich..directly in PC code..(without compilator), direct in machine code.
write this:
copy con ffs.com|ffs *comment None
176.230p144144230q195 *comment Hold "ALT" and write number
^Z *comment "CTRL+Z" then hit "Enter"
Explain:
copy(internal command)
con (command line switch)
^Z (CTRL+Z)
NOTE: Numbers in second line must be written with holdon "ALT" key and specified numbers.(ASCI Code). Non numbers symbols is normal keyboard selection. on input display should look like:
copy con ffs.com|ffs
░.цpРРцq├
^Z
Program is 9 bytes long... (simple but high effective)
Thats it.Reboot Computer... password is gone. CMOS is reseted to default values. Tested on Award 4.0,6.0, AMI BIOS. Break all..
and finaly again:
**To All readers of this thread: Please note that this thread is with regards to Self-Modding BIOS and are of NO Official Testings from MSI. MSI will not be responsible for ANY Damage caused from requesting these self modded BIOSes! Please mind that you'll ONLY request and flash AT YOUR OWN RISK!!!** also MODBIN Author(Borg N1) and me is dont are NOT responsible for ANY Damage caused by MODBIN or Add On Patch. (i didnt think there is risk exit must must said that warning....couse playing with BIOS is always risky). Happy Tweaking.

Well, I was wondering if the system not allowing the tty to change ownership -- but root also has the problem.
You might try checking which shell is set for your users in /etc/passwd, and then check that the shell really exists at the specified path, and that it is executable. You might also check that the home directory for you users (also specified in /etc/passwd) is mounted and your user has the appropriate rights to that directory.
Other than that, I am running out of altitude, airspeed and ideas all at the same time. There was one thing in your trace I did not understand. Why is your system looking to the floppy drive? I wonder if that is contributing to the problem.

Performance - starting to suspect underlying SAN

Oracle 11.2.0.1 Standard Edition One on Oracle Linux 5.6 - 64bit
non-RAC with ASM
Disclaimer - before taking this job about a month ago, I'd never had my hands on an ASM system. I'm also hobbled by the fact that we are not licensed for any peformance packs or grid control.
Starting two weeks ago we saw a dramatic increase in run times on several batch jobs. Jobs that normally took a minute started taking an hour, jobs that normally took an hour took all night. Nothing has changed in either the db config or the sql being executed. We have no base-line statistics from prior to the problem. I am beginning to wonder if there might be something going on with the underlying storage system - maybe something like a bad controller causing a lot of retries on I/O operations. Am wondering if there is any way for me to measure disk i/o peformance from ASM's point of view .. or even at the OS level. I've poked around a bit with iostat, but that's really ouside of my lane, and I'm not sure how ASM figures in.
From the DB side, we know we have some not-so-optimal SQL, but it was just as not-so-optimal before two weeks ago and still performing at an acceptable level. Yesterday we saw an update whose plan was driven by a FTS of an 8k row table take about an hour. Same process on a test server (where the db had just been replicated from production a couple of days earlier) took less than two minutes.
From where we (dba team) sit, the only thing that has changed was that a few days prior to the problem the SA un-mounted some disks that were NOT being used by the database. Part of that process requried a reboot of the server - presumably to rescan the available devices for mounting.

jgarry wrote:
(regarding the direct path) direct path reads and writes would show up in the statspack, as for example http://jonathanlewis.wordpress.com/statspack-distractions/ (which also is an example of how sometimes a top fiver is not on a critical path). Edit: Also see this thread: Re: serial table scan with direct path read compared to db file scattered read
What I'm after is all the configurations of the data path from Oracle to ASM. While ASM is in effect a raw disk, the underlying actual layout can have an effect, as for example the mentioned tiering. There can be so much more not reflected in what Oracle instruments. One thing is mount options - I'm guessing it is possible to dedicate a LUN for a raw device, but between all the Oracle options and bugs and linux options, get something odd happening with mount point aliases. You mentioned the SAN is not stressed - does that mean there is some linux buffer, some controller buffer, some SAN buffer serving requests. Could there be a problem with Oracle lying about how long the I/O waits are? The sreadtime statistics collection bug is probably what was tickling my brain, see http://hoopercharles.wordpress.com/2011/05/04/how-to-collect-statistics/ I can't help but wonder if you are using system statistics for the old hardware on the new, and kind of wonder if whoever came up with this system statistics feedback algorithm even contemplated hardware migration over time. Also an argument to patch up.
Edited by: jgarry on May 16, 2012 8:51 AMJoel,
Thanks for the additional comments. I have several things now I'm going to try and and collecting regular snapshots from statspack so I have a good baseline with which to compare results. I do feel that there is still a sense of "ready, fire, aim" because we've not been able to establish proof of a single cause - just a lot of theories. and for every theory there seems to be a piece of counter evidence.
I'm still working to understand the data path through ASM and the SAN components, but the SA told me yesterday the migration of the data from the EMC to the COMPELLENT happended nearly a year ago, so the idea that the move skewed or invalidated system stats (they've never been gathered, and that's on my to-do list) is ruled out as a proximate cause of recent heartburn.
And getting patched up was supposed to be my number-one priority until all this broke loose. At least my manager is very understanding and no one is asking for my head on a platter. I once worked at a company where if this had happend, I would have already been shown the door.

Get Axis range for 3D Ribbon Plot

When I plot something in a 3D Ribbon Plot with autoscale enable, the scale ranges (minimum and maximum) are automatically set in the proper way.
But after that, I can't read their values.
See the attached example: if I read Axis Range Maximum, I get 10 that is the value tha £D graph had before I plot the data.
Probably I'm missing something...
Vix
In claris non fit interpretatio
Using LV 2013 SP1 on Win 7 64bit
Using LV 8.2.1 on WinXP SP3
Using CVI 2012 SP1 on Win 7 64bit, WinXP and WinXP Embedded
Using CVI 6.0 on Win2k, WinXP and WinXP Embedded
Attachments:
3D Ribbon Plot.vi ‏17 KB

hanrock wrote:
I am the owner of the 3D Plots with the grey decoration. I subscribe the topics related these controls and collect bug/feedback to improve them in the future. Yeah, I create CARs for the bug I read from this forum. Sorry, I didn't standed up saying this is CARed or I cannot fix it for some reason. I dive more.
These controls are built by pure G code. So it has limitation if its sub-components control's property/event is not exported from VI server/event structure or XControl Framework does not support.
For the issue in this thread (range one), it can behave same as native XY Graph. For performance, do you mean it delays the interaction (rotate, click) or plot updates slow. I improved their performance somehow in LV2011. How big your input matrix is and what is the update frequency?
"... and there ws great rejoicing!" [Monty Python, after the minstrals death]
Well that is news to me.
I would like to invite you to post the CAR numbers so that we can track the bugs and fixes.
Simply posting those numbers will go a a long way toward changing my impression of the 3D graph issues. If you keep them secret we will be left to believe the issues are not being addressed, which has been the traditional viewpoint of users outside of NI. The Champions worked to change that attitude and to a large extent I feel the general outlook has been changing due to the public bug list that is no published for various LV versions.
The CAR numbers also give us a handle we can reach for when the apps we deliver to our customers screw up and we have to shrug our shoulders and tell them "When ever NI feels like it."
Trying to help more than I hurt and still fighting for the integrity of LabVIEW,
Ben
Ben Rayner
I am currently active on.. MainStream Preppers
Rayner's Ridge is under construction

COLLECT bug?

Similar Messages

Maybe you are looking for