Re: [U2] SB+ PA Compare
We use CVS, PreForce, Git, and Subversion within the Eclipse framework with our XLr8Editor. We never have to set directory type files that most version control systems require to understand dictionaries, procs, paragraphs, includes, and menus. We do not have to have any scripts or bat files to get and remove stuff from any version control system. You just right click on the item pick the Team menu option and load it into source control. It is that easy. Within Eclipse there is local tracking of changes. We sometimes keep 100 versions locally when we are busy changing a program or adding a feature or two. The compare editor is built-in to Eclipse and it comes up every time you try to edit something that is different from your local version. You can then merge those changes within the compare editor. You never have to leave Eclipse to do any of these functions. You can even tie all this into Bugzilla or many other bug tracking system that are free plug-ins for Eclipse. Why spend all this money in today's crazy environment that you can get this stuff for FREE or low cost. We are just finishing a feature that will keep your local version and database versions in sync based on what criteria you pick. This will be great add-on for keeping customers and all programmers abreast of changings happening in the U2 environment. Regards, Doug www.u2logic.com XLr8Editor is still 49.00 dollars, free trials available as well. ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
[U2] Unidata Write Error
We just moved to unidata 7.2.5 on RHEL 5 from unidata 7.1.8 on Solaris 9. In the conversion I ran convcode, convdata and convidx across the board. Generally speaking, things have generally gone smoothly but I just ran into an error which I can't track back to conversion, permissions or triggers, but I'm open to thoughts. When attempting to delete a record, here's what happens: :AE H08.WEB.USERS 11196 --: fd Record 11196 in file H08.WEB.USERS * You are about to DELETE the ENTIRE record from the file! OK? n/y y In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,89). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,303). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,102). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,169). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,210). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,17). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 [AE] UniBasic WRITE failed, STATUS=1, check triggers. If instead of going thru AE I attempt a DELETE I get this: :DELETE H08.WEB.USERS 11196 Index: invalid OV element: (34,89). Index: invalid OV element: (34,303). Index: invalid OV element: (34,102). Index: invalid OV element: (34,169). Index: invalid OV element: (34,210). Index: invalid OV element: (34,17). 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 '11196' is not a record id in H08.WEB.USERS I reran convdata to ensure that this file was processed and Unidata thinks it's fine. -- Jeff Butera, Ph.D. Manager of ERP Systems Hampshire College jbut...@hampshire.edu 413-559-5556 ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] Unidata Write Error
On 6/3/2010 1:47 PM, Jeffrey Butera wrote: We just moved to unidata 7.2.5 on RHEL 5 from unidata 7.1.8 on Solaris 9. In the conversion I ran convcode, convdata and convidx across the board. Generally speaking, things have generally gone smoothly but I just ran into an error which I can't track back to conversion, permissions or triggers, but I'm open to thoughts. When attempting to delete a record, here's what happens: :AE H08.WEB.USERS 11196 --: fd Record 11196 in file H08.WEB.USERS * You are about to DELETE the ENTIRE record from the file! OK? n/y y In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,89). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,303). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,102). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,169). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,210). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,17). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 [AE] UniBasic WRITE failed, STATUS=1, check triggers. If instead of going thru AE I attempt a DELETE I get this: :DELETE H08.WEB.USERS 11196 Index: invalid OV element: (34,89). Index: invalid OV element: (34,303). Index: invalid OV element: (34,102). Index: invalid OV element: (34,169). Index: invalid OV element: (34,210). Index: invalid OV element: (34,17). 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 '11196' is not a record id in H08.WEB.USERS I reran convdata to ensure that this file was processed and Unidata thinks it's fine. Do you have any alternate key index's built on this file. If you do, you will have a X_H08.WEB.USERS file. If you do, maybe either disable updating the index, or just rebuild it. If it is damaged, may have to delete the index and reindex. ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] Unidata Write Error
If the index on this file was heavily overflowed, it is possible that convidx didn't successfully convert the index. You need to: LIST.INDEX H08.WEB.USERS (to make sure you have all of the index names for recreating) DELETE.INDEX H08.WEB.USERS ALL (to completely delete the index file) CREATE.INDEX - indexes listed in LIST.INDEX BUILD.INDEX H08.WEB.USERS ALL (possibly using ONLINE keyword if you need to continue to update the primary file while the index builds) Regards, Wally Terhune U2 Support Architect Rocket Software 4700 S. Syracuse Street, Suite 400 **Denver, CO 80237 **USA Tel: +1.720.475.8055 Email: wterh...@rs.com Web: www.rocketsoftware.com/u2 -Original Message- From: u2-users-boun...@listserver.u2ug.org [mailto:u2-users-boun...@listserver.u2ug.org] On Behalf Of Jeffrey Butera Sent: Thursday, June 03, 2010 11:47 AM To: U2 Users List Subject: [U2] Unidata Write Error We just moved to unidata 7.2.5 on RHEL 5 from unidata 7.1.8 on Solaris 9. In the conversion I ran convcode, convdata and convidx across the board. Generally speaking, things have generally gone smoothly but I just ran into an error which I can't track back to conversion, permissions or triggers, but I'm open to thoughts. When attempting to delete a record, here's what happens: :AE H08.WEB.USERS 11196 --: fd Record 11196 in file H08.WEB.USERS * You are about to DELETE the ENTIRE record from the file! OK? n/y y In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,89). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,303). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,102). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,169). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,210). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,17). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 [AE] UniBasic WRITE failed, STATUS=1, check triggers. If instead of going thru AE I attempt a DELETE I get this: :DELETE H08.WEB.USERS 11196 Index: invalid OV element: (34,89). Index: invalid OV element: (34,303). Index: invalid OV element: (34,102). Index: invalid OV element: (34,169). Index: invalid OV element: (34,210). Index: invalid OV element: (34,17). 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 '11196' is not a record id in H08.WEB.USERS I reran convdata to ensure that this file was processed and Unidata thinks it's fine. -- Jeff Butera, Ph.D. Manager of ERP Systems Hampshire College jbut...@hampshire.edu 413-559-5556 ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] Unidata Write Error
On 06/03/10 13:54, Wally Terhune wrote: If the index on this file was heavily overflowed, it is possible that convidx didn't successfully convert the index Thanks Wally. Actually, in our conversion we ran convidx and also rebuilt indicies for ALL files. That said, after disabling both indicies and triggers for this file and recreating them, the problem has been resolved so I'll assume that the index file indeed was the culprit. -- Jeff Butera, Ph.D. Manager of ERP Systems Hampshire College jbut...@hampshire.edu 413-559-5556 ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] U2-Users Digest, Vol 14, Issue 3
On 6/3/2010 3:00 PM, u2-users-requ...@listserver.u2ug.org wrote: Message: 6 Date: Wed, 2 Jun 2010 17:30:53 -0700 From: Boydell, Stuartstuart.boyd...@spotless.com.au To: U2 Users Listu2-users@listserver.u2ug.org Subject: Re: [U2] SB+ PA Compare Message-ID: 7ee7790a950e094399b01520071826a20106b48d6...@sg1rd3xvs031.red003.local Content-Type: text/plain; charset=us-ascii Hi Steve, Perforce for source control combined with virtual machines for development and test. The use of Perforce was mainly because it has a great API and fits well with the spectrum of development. It's not all about U2 here. I haven't looked at any of the distributed VCS components so I can't offer any advice there. As a suggestion, I cannot recommend highly enough the use of VMs for dev and test. Using a VM will give you a system based replication of your production environment and data. This means rather than creating development and test accounts, you can develop and test in your production environment with all the same file pointers, account names, security settings etc. There are numerous other advantages to VMs. They can be provisioned and torn down again rapidly. Depending on your VM platform they can be snap-shot and rolled back during bug fixing, install dev/tests and scenario testing. You can have an app developed and installed with*exactly* the same set of artefacts and instructions for test and prod. And it also which means you can reduce the number of accounts you need to maintain. Cheers. S I would love to use vm's, however, we are using udt on aix. there isnt any guest ability for aix :). I could maybe use rhel test accts, but the conversion of files each time could be a problem. The endian problems, etc. For those Prelude users out there, I did hack up a PSICOPY that uses rsync so I can psicopy remotely. That still wont help with the conversion issues of mixing x86 and ppc. ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] SB+ PA Compare
On 6/3/2010 3:00 PM, u2-users-requ...@listserver.u2ug.org wrote: Message: 4 Date: Wed, 02 Jun 2010 21:59:17 +0100 From: Anthony Youngmanu...@thewolery.demon.co.uk To:u2-users@listserver.u2ug.org Subject: Re: [U2] SB+ PA Compare Message-ID:4c06c625.3090...@thewolery.demon.co.uk Content-Type: text/plain; charset=ISO-8859-1 I'm not a mercurial user - I'm learning git, but what's wrong with a DVCS? Yes it would be nice if each dev could have their own account, but you're assuming one dev per DVCS instance. Why not one account per DVCS instance? But by not giving each dev their own account, you're throwing all the advantages of ANY version control system! Whether it's Mercurial, git, SCCS, CVS, Visual Source Safe, as soon as you have devs fighting over access to the source you're giving yourself a headache. Centralised system, distributed system, it's all the same. Let me give you an example method of working with git, several devs, one test environment. Pretty safe bet you can translate this into Mercurial terms easy enough. Each dev has their own test account, probably can't do an awful lot there, but they do all their coding, compiling, and basic unit testing in their own account. They can then push their changes to a branch in the central test account to test them. When they're happy, they can merge them into the main branch. If several devs are fighting over the central test account, the fact that git allows very rapid changes between branches means that one can have it for however long it takes, then revert his changes leaving it clean for the next one, so you can switch the test account rapidly. And more importantly, dev A*can't* step all over My experience of cvs's commercially was either none (for UV, had to track everything manually), or VSS for VB where it was a right pain when I had something checked out for dev work and somebody else needed it for a quick bugfix, or vice versa. Thing is, you can put procedures in place so that a DVCS has strict central control. You can't set up a centralised VCS to give several developers flexibility to work on the same code at the same time. Cheers, Wol I think this is the best solution yet. Our data is ~15g/account, would be expensive, but might be best. ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] U2-Users Digest, Vol 14, Issue 3
IBM AIX supports WPARs, which allows us to run multiple instances of UniVerse on the same server. Perhaps UniData will work as well. Best regards, Henry Unger Hitech Systems, Inc. http://www.hitech.com -Original Message- From: u2-users-boun...@listserver.u2ug.org [mailto:u2-users-boun...@listserver.u2ug.org] On Behalf Of Steve Romanow Sent: Thursday, June 03, 2010 12:08 PM To: u2-users@listserver.u2ug.org Subject: Re: [U2] U2-Users Digest, Vol 14, Issue 3 On 6/3/2010 3:00 PM, u2-users-requ...@listserver.u2ug.org wrote: Message: 6 Date: Wed, 2 Jun 2010 17:30:53 -0700 From: Boydell, Stuartstuart.boyd...@spotless.com.au To: U2 Users Listu2-users@listserver.u2ug.org Subject: Re: [U2] SB+ PA Compare Message-ID: 7ee7790a950e094399b01520071826a20106b48d6...@sg1rd3xvs031.red003.local Content-Type: text/plain; charset=us-ascii Hi Steve, Perforce for source control combined with virtual machines for development and test. The use of Perforce was mainly because it has a great API and fits well with the spectrum of development. It's not all about U2 here. I haven't looked at any of the distributed VCS components so I can't offer any advice there. As a suggestion, I cannot recommend highly enough the use of VMs for dev and test. Using a VM will give you a system based replication of your production environment and data. This means rather than creating development and test accounts, you can develop and test in your production environment with all the same file pointers, account names, security settings etc. There are numerous other advantages to VMs. They can be provisioned and torn down again rapidly. Depending on your VM platform they can be snap-shot and rolled back during bug fixing, install dev/tests and scenario testing. You can have an app developed and installed with*exactly* the same set of artefacts and instructions for test and prod. And it also which means you can reduce the number of accounts you need to maintain. Cheers. S I would love to use vm's, however, we are using udt on aix. there isnt any guest ability for aix :). I could maybe use rhel test accts, but the conversion of files each time could be a problem. The endian problems, etc. For those Prelude users out there, I did hack up a PSICOPY that uses rsync so I can psicopy remotely. That still wont help with the conversion issues of mixing x86 and ppc. ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] wIntegrate equivalent of System BuilderTU.CLIENT.GETENV subroutine
Gregor,
Thanks so much - I was still confused as I never used script before,
till someone pointed out the example - I guess I missed that when I read
your response the first time.
Works great!
Thanks!
Doug
-Original Message-
From: u2-users-boun...@listserver.u2ug.org
[mailto:u2-users-boun...@listserver.u2ug.org] On Behalf Of Gregor Scott
Sent: Sunday, May 30, 2010 6:44 PM
To: 'U2 Users List'
Subject: Re: [U2] wIntegrate equivalent of System
BuilderTU.CLIENT.GETENV subroutine
Doug,
there is a client-side function called GetEnv() which you can use:
From the 6.1.3 Client scripting manual--
Syntax
GetEnv var_name, [check_exists]
Description
The GetEnv function gets the value of an OS environmental
variable.
If the variable
does not exist it returns . The variable can be checked to see
if
it does exist by
passing true as a second parameter to the function.
Parameters
The following table describes the parameters of the GetEnv
command:
Parameter Description
var_nameThe name of the variable enclosed in
quotes.
check_existsIf specified as true this parameter will return
true
(1) if the variable exists or false (0)
if
it doesn't.
Return Value
The value of the environmental variable unless check_exists is
true
in which case
either true or false is returned.
Example
Display the TEMP variable on Windows
MessageBox The TEMP variable is :GetEnv(TEMP)
Version
6.0.3 Original version
-
The host subroutine WIN.EVAL can be used to return the value of an
environment variable:
call WIN.EVAL('GetEnv(TEMP)',val.TEMP)
Gregor
-Original Message-
From: u2-users-boun...@listserver.u2ug.org
[mailto:u2-users-boun...@listserver.u2ug.org] On Behalf Of Doug Farmer
Sent: Saturday, 29 May 2010 12:52 AM
To: U2 Users List
Subject: [U2] wIntegrate equivalent of System Builder TU.CLIENT.GETENV
subroutine
For all of you multiple platform developers - A query:
I have need to duplicate some code I wrote in System Builder in a
wIntegrate environment.
System Builder has a subroutine called TU.CLIENT.GETENV. I use this
subroutine to look up an environment variable on a user's PC so I know
where a program I need to execute is loaded. I am not able to find an
equivalent command using wIntegrate subroutines.
Not being a wIntegrate programmer, I resorted to reading all 600 pages
of the wIntegrate Host Subroutines Reference manual. Quite a barn
burner! I could not put it down.
But seriously, I need a way to look at the user's environment settings
and pull the value for a specific setting.
For example, under System Builder, I would have this code
CALL TU.CLIENT.GETENV(CCPath,ProgramPath)
Command = '1':ProgramPath:'\CCEntry.exe'
CALL TU.EXECUTE.SHELL(Command,'')
Under wIntegrate, I found the WIN.PCRUN subroutine (or WIN.PCRUNS2), so
I know I can run the program, but I do not know how to get the
environment setting or where the program is installed.
Any help is greatly appreciated!
Doug Farmer
This e-mail is for the use of the intended recipient(s) only. If you have
received this e-mail in error, please notify the sender immediately and then
delete it. If you are not the intended recipient, you must not use, disclose or
distribute this e-mail without the author's prior permission. We have taken
precautions to minimize the risk of transmitting software viruses, but we
advise you to carry out your own virus checks on any attachment to this
message. We cannot accept liability for any loss or damage caused by software
viruses. Any views and/or opinions expressed in this e-mail are of the author
only and do not represent the views of Epicor Software Corporation or any other
company within its group.
This message has been scanned for malware by Websense. www.websense.com
___
U2-Users mailing list
U2-Users@listserver.u2ug.org
http://listserver.u2ug.org/mailman/listinfo/u2-users
[U2] Filesystem options cio for AIX 5.3
I am working with an AIX 5.3 system using Unidata 7.1.10 We are having some performance issues that are directly related to I/O. The files have been resized and brought down some of the I/O problems. In the past with some Oracle system I have had to administer ( I still prefer the mv world ), we used the cio options for mounting filesystems which allows for concurrent reads and writes. Has anyone tried this with Unidata or know of any pitfalls to doing so. Thanks in advance -Dan ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
[U2] Network VOC pointer
I had two Win 2K3 machines running in our development environment. I created a VOC pointer within UD on one server to a directory in another server. e.g. DWHELP 001 DIR 002 \\WEBSERVER\E$\MYCOMPANY\HELP 003 D_VOC This worked perfectly and allowed updating of help items from within UD. I've installed Windows 2008 R2 onto the development web server (not the UD server, which still runs 2K3), and set up the environment. I shared the E:\WevDev\MyCompany\Help directory for everyone. I've turned off the 2K8 passwork protected sharing in Advanced sharing settings and I'm able to access the directory from the dbms server in Windows explorer by entering \\webserver\help. No credentials are requested or anything. However, when I change the VOC entry too... 001 DIR 002 \\WEBSERVER\HELP 003 D_VOC ...and try to list the directory contents it fails in UD. 3 Dev (0)- LIST DWHELP errno=2: No such file or directory can not stat() in U_get_fileid(),fname=\\WEBSERVER\HELP. Open file error. 3 Dev (0)- Does anyone have an idea why this doesn't work? Thanks, Bill ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] Filesystem options cio for AIX 5.3
Don't know about cio, but if you do a vmstat -v | grep buf today, and then another one in 24 hours, are you seeing the numbers climb? From: dgo...@interlinebrands.com To: u2-users@listserver.u2ug.org Date: Thu, 3 Jun 2010 16:51:19 -0400 Subject: [U2] Filesystem options cio for AIX 5.3 I am working with an AIX 5.3 system using Unidata 7.1.10 We are having some performance issues that are directly related to I/O. The files have been resized and brought down some of the I/O problems. In the past with some Oracle system I have had to administer ( I still prefer the mv world ), we used the cio options for mounting filesystems which allows for concurrent reads and writes. Has anyone tried this with Unidata or know of any pitfalls to doing so. Thanks in advance -Dan ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users _ Hotmail is redefining busy with tools for the New Busy. Get more from your inbox. http://www.windowslive.com/campaign/thenewbusy?ocid=PID28326::T:WLMTAGL:ON:WL:en-US:WM_HMP:042010_2 ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] Filesystem options cio for AIX 5.3
Dan, We have done a lot of work on tuning AIX for UniVerse. I looked into CIO and its potential benefits to UV. I even raised a case with U2 support. Their response was that as CIO essentially eliminates caching updates in memory it requires the update to be physically made to disk before the process can continue. The nature of UV (and possibly UD) updates made U2 support think there would not be much benefit. The biggest improvements we made were to how AIX manages memory and its I/O buffers. The memory changes made a huge different to process paging, and the I/O buffers played a large part in stabilising our I/O activities. These settings are the culmination of taking advised adjustments from various sources and then fine tuning for our UV installations. ___Memory Management Parameter Default New Value Command to Adjustment minperm%20 5 vmo -p -o minperm%=5 maxperm%80 90 vmo -p -o maxperm%=90 maxclient% 80 90 vmo -p -o maxclient%=90 lru_file_repage 1 0 vmo -p -o lru_file_repage=0 page_steal_method 0 1 vmo -p -o page_steal_method=1 ___PBUFS___ pv_min_pbuf 512 4096 ioo -p -o pv_min_pbuf=4096 ___JFS2___ j2_dynamicBufferPreallocation 16 128 ioo -p -o j2_dynamicBufferPreallocation=128 j2_nBufferPerPagerDevice512 8192ioo -p -o j2_nBufferPerPagerDevice=8192 ___JFS___ numfsbufs 196 2048 ioo -p -o numfsbufs=2048 maxpgahead 8 16 ioo -p -o maxpgahead=16 I found the following links very helpful when exploring the AIX performance tuning options: http://www.circle4.com/blosxomjl.cgi/ http://users.ca.astound.net/baspence/AIXtip/j2_buffer.htm This PDF explains the benefits of tuning Memory: http://www.ibm.com/developerworks/wikis/download/attachments/53871915/VMM+Tu ning+Tip+-+Proctecting+Comp+Memory.pdf?version=2 Whilst focused on Oracle, this PDF provides an interesting snippet on page 9: Large hdisks are evil for I/O performance! http://www.circle4.com/papers/PAI22.pdf Hope this helps Gregor -Original Message- From: u2-users-boun...@listserver.u2ug.org [mailto:u2-users-boun...@listserver.u2ug.org] On Behalf Of Dan Goble Sent: Friday, 4 June 2010 6:51 AM To: U2 Users List Subject: [U2] Filesystem options cio for AIX 5.3 I am working with an AIX 5.3 system using Unidata 7.1.10 We are having some performance issues that are directly related to I/O. The files have been resized and brought down some of the I/O problems. In the past with some Oracle system I have had to administer ( I still prefer the mv world ), we used the cio options for mounting filesystems which allows for concurrent reads and writes. Has anyone tried this with Unidata or know of any pitfalls to doing so. Thanks in advance -Dan ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users -- Message protected by DealerGuard: e-mail anti-virus, anti-spam and content filtering. http://www.pentanasolutions.com Click here to report this message as spam: https://login.mailguard.com.au/report/1A12oPLg1S/6aKdA6wbF9XEaIVKtBO7mB/3.05 3 ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] SB+ PA Compare
Yep - we use AIX snapshot, uvbackup / uvrestore to rhel. Solves endian problems. Then the only issue is fixing files with triggers (fix is scripted but it's a pain in the bum) -Original Message- I would love to use vm's, however, we are using udt on aix. there isnt any guest ability for aix :). I could maybe use rhel test accts, but the conversion of files each time could be a problem. The endian problems, etc. ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
Re: [U2] Unidata Write Error
Hi Jeff, When we made the move to RHEL (while I was @ Beloit College), we had something similar happen. I had to do more than rebuild the indexes. I had to completely delete the indexes, recreate them, and than rebuild them. I'm pretty sure I only had to do this for some files. I do remember it being a pain in the neck. I believe Datatel had an AnswerNet Doc on the subject. Cheers, Jon Wells jwe...@campbell.edu At 01:47 PM 6/3/2010, you wrote: We just moved to unidata 7.2.5 on RHEL 5 from unidata 7.1.8 on Solaris 9. In the conversion I ran convcode, convdata and convidx across the board. Generally speaking, things have generally gone smoothly but I just ran into an error which I can't track back to conversion, permissions or triggers, but I'm open to thoughts. When attempting to delete a record, here's what happens: :AE H08.WEB.USERS 11196 --: fd Record 11196 in file H08.WEB.USERS * You are about to DELETE the ENTIRE record from the file! OK? n/y y In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,89). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,303). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,102). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,169). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,210). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 Index: invalid OV element: (34,17). In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 In /usr/ud72/sys/CTLG/a/AE_AE at line 2285 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 [AE] UniBasic WRITE failed, STATUS=1, check triggers. If instead of going thru AE I attempt a DELETE I get this: :DELETE H08.WEB.USERS 11196 Index: invalid OV element: (34,89). Index: invalid OV element: (34,303). Index: invalid OV element: (34,102). Index: invalid OV element: (34,169). Index: invalid OV element: (34,210). Index: invalid OV element: (34,17). 311296:block offset error in U_blkwrite for file 'H08.WEB.USERS', key '', number=131 1:block offset error in delete_from_group for file 'H08.WEB.USERS', key '11196', number=131 '11196' is not a record id in H08.WEB.USERS I reran convdata to ensure that this file was processed and Unidata thinks it's fine. -- Jeff Butera, Ph.D. Manager of ERP Systems Hampshire College jbut...@hampshire.edu 413-559-5556 ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users ___ U2-Users mailing list U2-Users@listserver.u2ug.org http://listserver.u2ug.org/mailman/listinfo/u2-users
