|
We converted from MAPICS I to BPCS about 15 years ago & I have been with the same company that long. If you are not already on the MAPICS_L here at midrange dot com, I suggest you join since they probably have guidance on managing MAPICS files. You might also find information of value in the archives of that group. I do remember that the MAPICS documentation was first rate, compared to BPCS. We had all 3 sets ... the behind the scenes IT staff manuals that explained what MAPICS was doing & the layouts of various files & explanations of what all those codes were for in various fields ... the general management manuals that showed what needed to be done in each application for it to function properly ... there was one of them for IT system support ... then there were the run manuals for the people doing the hands on each application. There was an expense associated with getting this documentation & generally there were certain manuals that were VERY popular in great demand & 90% of them sat gathering dust ... I would see if this stuff is available today on like CD Rom. One big problem I had was in staying current documenting modifications. The MAPICS manuals were loose leaf notebooks & we were inserting pages with information on what we had changed. I could handle this using SEU but management wanted something that looked professional, and I did not have direct access to word processing at the office, so it was either do it at home which I did not want to do, or use office people who were hopelessly back logged. Many good ideas here in MIDRANGE_L posts ... we only got 12G of space & ours been around 70-85% for the last year ... I am slowly bringing it down overall by analysing what dead records are in our ERP that we do not need ... it was over 100% a few years ago & that is a real bad trip you not want to take. Do you have end users creating queries or some other method of reports & inquiries on the fly? We do create some work files with query. Some of our users are inexperienced in data base design & could create excessive query work file collections. Have you had some outside programmers working your place in the time period of the growth? I find that such people have their own standard practices & can leave a trail of libraries & copies of software in excess of what I would consider reasonable. We use GO CLEAN to automate some clean-ups. I set message queues to save stuff for 10 days ... like a week plus a weekend. I tried to save system stuff for less than 30 days but it would not let me, so I go in regularly to wipe it all out ... we do not have any system logs or journals that are more than a few days old, except when there is a known problem to investigate. We also do IPL about once a month ... I believe the IPL does reclaim disk space of deleted spool files & this explains to me why each IPL does recover about 3% disk space that is gone again in a few days. Prior posts here warned that some of these options can take a while to run, so you might want to search the archives for each RCL suggested to see if there are any gotchas. I monitored WRKJOBTBL for a few months to see about adjusting what setings advisable, but we now have our system on automatic tuning. We encourage people to save their print outs until after backups, because if something goes wrong before backup they may need to recover from audit trails what they did. For some applications we encourage people to save their print outs for a week because it may take that long before it is evident that something went wrong. But beyond that, we remind people who have print-outs much longer, that we are not backing up print-outs & they should delete those that they do not need because large numbers of large print-outs do eat disk space & jobs performance. The accounting department used to keep their month end reports on-line for months ... I suggested to them that from a navigation standpoint it might be useful to move 100% of those for a particular end-month to an output queue just for that end-month, and if they really really needed this stuff on-line which we are NOT backing up, they should authorize purchase of a spool archiving deal. I say they used to do this. They don't any more. There was a dramatic drop in people saving old spool files after we showed them how to transfer AS/400 reports to PC files that could be e-mailed or into Excel or Word document. After Backup, I check the Job Dumps to see if there were any on a USER, as opposed to on some IBM stuff & those on a USER I transfer to a different spool, then I clear the entire JobDump Spool ... I do not get around to this every nite. If you not been doing this, WRKOUTQ gets you at list of all of them, then scroll down & only view the Qs with significant numbers of contents. I use dumps to *OUTFILE pretty heavily now compared to a year or so ago when I first learned on midrange_L & AS/400 network how to work them. One of them runs off the job scheduler several days a week to tell me how many bytes we have of deleted records, and when the number climbs to say 5+ Meg that trips me into doing a general reorg that evening. When I first started doing this, we had over 500+ Meg in this situation. I have another one that runs off the job scheduler to populate a work file of the SIZES of our files in selected libraries. It does this Monday 4 am ... then another one is populated Friday 4 pm followed by a COMPARISON of the two work files to give me some reports showing which files are getting the most significant growth & other changes in sizes. If you are not familiar with the IBM job scheduler ... GO CMDSCDE be careful about putting any MAPICS jobs on this thing When a user signs on, in our case BPCS, it sets up library list & a bunch of conditions needed for menu options to work correctly. When something runs off the job scheduler, there is no library list or all that other stuff native to it, so to get your ERP stuff running off the job scheduler, needs a programmer who has a keen understanding of what is going on under the covers. Our retrieve disk info is set to run very infrequently since I am now finding that my outfile analysis is much more useful in analysing what files in most need of cleanup. I have an RPG program that reads an outfile of our files directory, that prints a report on how many members it has that are empty & how many populated, numbers of records there, for the big ones an SQL COUNT(*) those older than a year & incidentally what is the oldest date in this file ... you have to know file layout to do that. I have been using this tool for the last 6 months to prioritize what files in most need of old deadwood clean out. I have found where there are members with zero records that have not been accessed in over a year ... I have plans to say goodbye to them that I have not actually implemented yet. There is also an issue associated with those files that grew to millions of deleted records before I figured that out & reorged them ... they may actually be the capacity they were before I started regular reorg ... I may need to downsize some. Depending on your software package, you may have a month / year cycle in which data grows to some size in some files, then gets cleared away at the end of end-month or end-year processing. For example, we keep our inventory transactions history on-line for 365 days. When we do our physical inventory, that's several thousand transactions added in one day. Now there are alternative ways to do physical inventory that could add hundreds of thousands of transactions. We have people who are trying to figure out how to run the BPCS cycle counting system. I know that because I run the INV900 during EOM because that task is fraught with risks of going awry & its first screen invariably tells us that someone tried to do cycle counts this month & did not get the task to a conclusion. I have a suspicion this adds records some place, but I have not yet taken the trouble to track it down. Our General Ledger is currently done in summary format. A new person in our accounting department changed it to detail format, when noodling around without reading the documentation. I found out about it when they ran out of Journal Batches before the end of the month. Problem was solved, but here a decision by someone who did not understand their department application led to a short term dramatic rise in the volume of their transactions. You have added 30% consumption in 2 months ... did you install some package 3 months ago & did it come with some system for estimating disk space needs? Did the disk space consumption grow steadily, or did you check this intermittently & suddenly notice it had jumped & you don't know if it came all at one time. I try to remember to check this story before & after implementing some new application so I know what the impact is. There are tons of computer magazines. I find that News/400 recently rebranded iSeries News to be the BEST & their iSeriesNetwork.com to be a dynamite complement to midrange dot com, but there is a cost to subscribe. We also get AS/400 magazine also rebranded iSeries Magazine which is FREE from IBM. This latter has decent tips & also good stuff on their web site ... you might check the general resources area off of this midrange dot com main page for other links. We get several BPCS newsletters ... there are probably some MAPICS magazines out there. You mentioned college classes. You might try to schedule a trip to COMMON or IBM technical conference. IBM has great classes & you can take some of them on-line. http://www.iseries.ibm.com/developer/education/ibo/view.html?all IBM offers classes all over the USA ... check out if any of these cities are near you http://www-3.ibm.com/services/learning/lsweb/educationcenters/ As for which classes would be of greatest value for you to take http://www-3.ibm.com/services/learning/community/as400/ If you going to be doing several IBM classes http://www-3.ibm.com/services/learning/us/ There also used to be a MAPICS UNIVERSITY If you are any place near Evansville Indiana, our local *BASE user group intermittently brings some IBM instructor to teach a class here to folks from many organizations in our user group. The AS/400 user group circulates surveys to the membership to find out what classes there is an interest in. > Subj: RE: Oh where has my disk space gone? > Date: 11/15/2001 9:41:35 AM Central Standard Time > From: rsnyder@atlasdie.com (Rebecca Snyder) > Sender: midrange-l-admin@midrange.com > Reply-to: <A HREF="mailto:midrange-l@midrange.com"> midrange-l@midrange.com</A> > To: midrange-l@midrange.com > > Thanks everyone for all the suggestions. I've set RTVDSKINF to > run tonight at 8pm. I've never run it before so I won't have > anything to compare it to, but maybe something will stick out. > > We only save 4 days of performance data, which when I deleted > 3 days worth it got me .1% of space back. > > We do run cleanup and keep the following: > > Number of days to keep: > User messages . . . . . . . . . . . . . . . . . 15 1-366, > *KEEP > System and workstation messages . . . . . . . . 15 1-366, > *KEEP > Job logs and other system output . . . . . . . 30 1-366, > *KEEP > System journals and system logs . . . . . . . . 30 1-366, > *KEEP > OfficeVision for AS/400 calendar items . . . . 0 1-366, > *KEEP > (we don't use office vision) > > Are we saving too much? What do you suggest for these options? > > We also do a monthly IPL. In fact we just did one this Saturday. > I've never run RCLRSC or RCLSTG for that matter. We've had our > AS/400 (iSeries) since June 1999, and this is our first one. > We only use Mapics for our carbon paper business. Our steel rule > die business still runs our home-grown system on our VAX. Our > computer room did not implode when we put an AS/400 next to our > VAX!! I'm planning on doing a RCLSTG on some weekend in December. > Would it be better to do RCLRSC or RCLSTG? Or maybe both? > > So needless to say I was kind of thrown on the thing without any > training (except in mapics) and everything I've learned over the > past 2 1/2 years has been out of necessity or from reading > the messages posted here. I did manage to get copies of "Control > Language Programming for the AS/400" and "Mastering the AS/400", > both which has been very helpful although I haven't had the time > to read through either of them. I've just picked out what I've > needed at the time I needed it. One of these days I'm going to > see if one of our local colleges has classes that I can take > after hours... > > I'll stop rambling now & thanks again for all the help! > > Rebecca Snyder > > > > -----Original Message----- > > From: Rebecca Snyder [SMTP:rsnyder@atlasdie.com] > > Sent: Thursday, November 15, 2001 4:16 PM > > To: midrange-l@midrange.com > > Subject: Oh where has my disk space gone? > > > > In the past 2 months my disk space has gone from > > 45% used to 75% used... and I've got 50G of space! > > > > How can I tell what's using up my disk space? Because > > we couldn't possibly be selling that much carbon paper! > > > > > > Rebecca Snyder > > Atlas Companies, Inc > > 219-295-0050 ext 242 > > mailto:rsnyder@atlasdie.com MacWheel99@aol.com (Alister Wm Macintyre) (Al Mac) BPCS 405 CD Manager / Programmer @ Global Wire Technologies Incorporated http://www.globalwiretechnologies.com = new name same quality wire engineering company: fax # 812-424-6838 Sep 11 Favorite Links: http://www.nzherald.co.nz/pdf/middle_east.pdf http://www.semitrue.com/thankyou/ http://groups.yahoo.com/group/TYR http://www.skirsch.com/politics/plane/disable.htm http://www.geocities.com/wasabidoh/Pictures.html - select Attack on America Newspapers World Wide http://www.wheretodoresearch.com/news/foreign_newspapers.htm http://www.wheretodoresearch.com/news/US_Newspapers.htm Intelligence Briefings by country http://www.nsdmg.org - click on REAL WORLD RESOURCES http://www.c-span.org/international/links.asp http://www.cnn.com/2001/WORLD/asiapcf/central/09/17/asia.support/ http://www.odci.gov/cia/publications/factbook/geos/af.html http://www.economist.com/countries http://www.washingtonpost.com/wp-dyn/world/search/list/index.html http://www.debka.com/ http://www.stratfor.com
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.