Enter the access code after executing PAEZIP/AUTHORIZE. If you do not have an access code, call PAE at 978-744-8612 to obtain an access code.
Most of the links in this manual require that you have an internet connection. Clicking on the clink will go to WWW.PAEINC.COM.
The most current documentation will be the help text itself. As PAE gets input from users, we will change the product. Updates to the product will be made available at our web site WWW.PAEINC.COM. Check the Download section for the last update date.
In order to determine the objects that can be archived, the PAEZIP/ARCHIVERPT OPTION(*PRINT) should be submitted to a batch subsystem. PAE does not recommend that this command be executed interactively since it will take a while to process. Basically, this command goes out and looks at all the objects on the system. It determines which objects will be placed on the archive report based on the addition parms (days/size) in the command. These parameters may be modified from the command defaults. It is recommended that you leave the defaults on the first pass of the command. After a report has printed, look at the bottom of the report to see how much storage can be saved. You want to change the number of days and object size to see how that effects the total amount of storage that can be saved. After determining the number of days and object size, you are now ready to load the Archive File. Loading the file is done with the same command but the OPTION(*FILE) is used instead of *PRINT. This command should be added to the IBM free job scheduling system with the ADDJOBSCDE command. See below.
You can change the execution time and day to meet your companies priorities. Before you leave that evening you should do a WRKJOBSCDE PAELODARC. This will let you execute the command immediately by placing a 10 next to the job and pressing enter. This will provide you with the SBMJOB prompted. Press F10 to get additional parms.
Roll until you see the screen below. Change the execution date and time that you want the command to execute for the first time. Remember, the this method will only change it for this execution.
After the job finishes, review the report that is produced. Run the report again with different number of days and object size and see what the results are on the report. See a sample of a report below:
Pressing F10 on the Object Archive Management panel will display the current system defaults. These are the defaults that the command PAEZIP/ARCVHIVERPT will use when it is executed unless the command is changed. See panel below.
Minimum days is the number of day since the object was lasted used. If a 90 is the default, this indicates that all objects that have not been used in 90+ days will be a candidate for archiving. Minimum size is the size of the object in bytes. In the example below, an object would not have been touched/accessed in 90+ days and have an object size of 1,000,000 bytes or more to be added to the archive file.
"Y" means that the command PAEZIP/ARCHIVE OPTION(*COMP) is going to be executed in the IBM free job scheduling system. To view this job execute, WRKJOBSCDE PAEARCHIVE to see when it is going to execute. You can adjust the execution date and time schedule to meet your companies needs. The shipping default for this is a "N". It is IMPORTANT that you change this to a "Y" so that ISM will automatically place a job in the job scheduling system. The archive job in the scheduling system should be done AFTER the save of the PAEZIP directory and library.
In order for a particular object to be archived, it has to have been saved within that time number of days. If object was saved this morning, than the object would be archived this evening. If object has not been saved in a month, then QSYSOPR message queue would be notified. QSYSOPR would be given the opportunity to save the object immediately to a SAVF or a Device. QSYSOPR can bypass the archiving of the object. It is the users responsibility to save critical data before that data is archived. This would be extremely important if the object were a source file that was in constant update mode. While the 400 is extremely reliable, there are times when mistakes happen. It is better to be protected. If an object that is not compressed and has not been saved in X number of days (Maximum days since save) then and *NOT SAVED* will appear on the Archive file panel. It is important that an object be saved before compressing it.
The control fields on this panel are:
User Option:
*AUTO will automatically decompress object if touched/accessed, *USER indicates that the user will be asked if the object should be decompressed. It will be the users decision to decompress or not to decompress. The person who touched/accessed the object is the one getting the message.
CMP:
Compression is N = normal or X = maximum. Normal compression does not take too long but the compression is not dense. Maximum compression is slower but the compression is more dense, thus saving more space. If Member processing is desired, X=maximum compress is mandatory. Normal compression will not work for member processing. Journal receivers cannot be processed with X=maximum compression. Journal receivers have to be compress using N = Normal compression. You will get a error message when trying to compress a receiver with maximum compression.
IBM libraries starting with a Q*, for example QUSRSYS, have to use N = Normal compression. Q libraries cannot use X = maximum compression.
HLD:
Hold determines if an object should be compressed. N = object will be compressed or Y = object is held and will not be compressed until changed to a N. Objects that are placed into the archive file by the PAEZIP/ARCHIVERPT (*FILE) will automatically be given a Y indicating that the object will NOT be compressed until it is changed to N. This gives you the ability to control when a particular object is to be compressed. User Option, above, controls decompression. **IMPORTANT**
MBR:
Members in a file are to be processed individually. This in important when source files like QCLSRC is to be archived on a member basis. Member archiving of source files is ideal place to use this control field. Source files take up a lot of space. Usually, only 2-3% of a source file is used at any time. Therefore, you are wasting 97-98% of disk space.
DLY DAYS:
Number of days an object is not to be touched/accessed by the compression routine. 5 = do not touch the object for the next 5 days. After that, the object can be compressed. For example, let say that you were working an accounts payable program in QRPGSRC. It would not no sense to constantly compress and decompress this member daily. Determine how long that you would be working on this RPG member. Let's say that it would take you a week to complete the changes to the program, then change DLY DAYS to 7 or 8. That will indicate that the member will not be compressed or decompressed for that number of days.
The first compress should be done when there are plenty of system resources available. The amount of time to do the first compression is unknown. It depends of the processor on your system and the amount of memory in the system. It could take a considerable amount of time to compress all of the objects the FIRST TIME. Therefore, PAE is recommending that the compression take place in stages over a 4 days. Since a default of HLD = "Y" is placed on an object when is first placed into the archive file, no objects will be compressed until the Y is changed to a N in the control field HLD. Place an "N" next to those objects you want to compress. There might be objects that you do not want to compress. Leave those objects with HLD=Y. If you have 400 objects that are to be compressed, place a "Y" next to only 100 of those objects. If the first 100 finish quickly, then do the next 100. By staggering the compression, you will have better control of compress execution time.
In the panel below, QGPL/QCLSRC has been added by hand to the archive file. Also the control fields were changed. Since member processing is desired, a Y=MBR. Because of member processing, an X=CMP is used because N cannot be used if member processing is to take place. Most source files would NOT be automatically be placed in the archive file. The number of days since last used would most likely be less than the number of days in the system defaults. See system defaults above for more information
Some files such as source files will not meet the criteria of the SYSTEM DEFAULTS. Because of that these files will have to be added by hand, see above Adding Object by Hand. Usually these files might be volatile because of changes to the member in the file. Therefore, it is recommend that these files be saved to tape when they are in a decompress state. Follow the rules on member processing. Usually, only 3-5% of any source file is active. It is a total waste of space for an entire source file to be keep decompressed when only 3-5% of any source file is active.
MBR = Members are to be processed individually. This in important when source files like QCLSRC are to be archived on a member basis. Member archiving is ideal place to use the control field. Member decompress will not take very long because of the size of the member.
There might be times when you just want to decompress an object without touching with some program or query. Simply enter a *D or *DECOMP in the library name field next to the object to be decompressed. The object will be dynamically/automatically be decompressed.
Another method of starting decompression is the execute the PAEZIP/ARCHIVE OPTION(*DECOMP) command. This should be done in batch and when there are amply system resources available. You can enter a specific library/object or leave the default of *all. All objects in the archive file will be compressed that are not currently decompressed.
There might be times when you just want to compress an object without touching with some program or query. Simply enter a *C or *COMP in the library name field next to the object to be decompressed. The object will be dynamically/automatically be decompressed.
There might be times when you just want to compress all decompressed objects. Simply press the F24 key as indicated below. Notice, a WARNING window popped up, answer the prompt as desired.
Another method of starting compression is the execute the PAEZIP/ARCHIVE OPTION(*COMP) command. This should be done in batch and when there are amply system resources available. You can enter a specific library/object or leave the default of *all. All objects in the archive file will be compressed that are not currently compressed. This is the best method to have the compression take place in QBATCH.
It is recommend that the objects that are to be archived should be saved to tape before the object is archived. This is an important consideration. It has always been our policy to save any object before we make major changes to the object, just in case. You can execute the SAVOBJ, SAVLIB or NONSYS command before starting up the archiving process. Most likely a NONSYS should be done before the first archive process takes place.
If there are objects that are changed on a regular basis, it is recommend that these objects be saved first before being archived. For example, QCLSRC is Compressed then Decompressed based upon when the object was accessed or touched/accessed. This would be true of objects like source file (QCLSRC, QRPGSRC or month end job files etc.). See Maximum Days Since Save above.
Depending on the amount of objects archived and the frequency of times objects are archived, save the directory PAEZIP which is located on the IFS on are regular basis. This should be done at a minimum of once a week. Below is the command used to save PAEZIP. The name of the device below depends on the names of your tape device(s). Modify accordingly. Add the command to your weekly saves that are performed. The library PAEZIP should also be saved at the same time. Do a WRKJOBSCDE PAEARCHIVE to see when archiving is to be run. Saving of the library and directory PAEZIP should be done BEFORE PAESRCVHIVE job is executed.
Depending on the amount of objects archived and the frequency of times objects are archived, save the library PAEZIP on a regular basis. It is recommended that a SAVLIB of PAEZIP be done when your weekly saves are performed. It is recommended that a SAVCHGOBJ OBJ(*ALL) LIB( PAEZIP) be done when your nightly saves are performed. The directory PAEZIP should also be saved at the same time. Your nightly saves should be done before
Compression is slower than decompression. Decompress will take about 55% percent less time as compression. Therefore, we recommend that the compression be scheduled to run when there are enough system resources available. The size of the object in bytes and the type of compression, as indicated below, will determine the time of compression and decompression. Another factor in compression/decompression is the amount of memory and the processor feature code on the 400. The more memory and the more powerful the processor (CPU), the shorter compression/decompression.
CMP
Compression is N = normal or X = maximum. Normal compression does not take to long but the compression is not dense. Maximum compression is slower but the compression is more dense, thus saving more space. If Member processing is done, X = maximum compress is mandatory. Normal compression will not work for member processing.
IBM libraries starting with a Q*, for example QUSRSYS, have to use N = Normal compression. Q libraries cannot use X = maximum compression.
There most likely are objects that you do not want to be compressed even if PAEZIP/ARCHIVERPT OPTION(*FILE) placed the objects in the archive file. There are two choices: a.) leave a "Y" under the HLD control field for that object. b.) Blanking out the library/object line which will remove the object from the archive file.
Journal receivers will NOT automatically be decompress like files when they are touched/accessed. However, receivers take a tremendous amount of space. PAE decided to archive journal receivers because many companies keep old receivers on their systems just in case. It is the users responsibility to tell Intelligent Storage Management System (ISM) to decompress the receiver. Enter the command PAEZIP/ARCHIVE. Find the receiver that you want to decompress. Place a *D or *Decompress in the library name field next to the receiver name.
There might be times when you just want to find out what is in the archive file. Execute the PAEZIP/ARCHIVE OPTION(*DSP) command. This will display the archive subfile showing the current status of all objects in the archive file.
There are times when you want to decompress a object because you need the object in it's regular state. Therefore you would execute the command PAEZIP/ARCHIVE OPTION(*DECOMP) OBJECT(LIB/OBJNAME). This will decompress the object. Now process the file as desired. When finished, execute PAEZIP/ARCHIVE OPTION(*COMP) OBJECT(LIB/OBJNAME). This will compress the object back. For example, look at the program below.
PGM
PAEZIP/ARCHIVE OPTION(*DECOMP) OBJECT(LIB/OBJNAME) /* decompress
object */
CALL PGM(ABCD) /* this program will process the object "objectname
in the library lib" */
PAEZIP/ARCHIVE OPTION(*COMP) OBJECT(LIB/OBJNAME) /* recompresses
the object */
ENDPGM
There are times when you might need a very large object decompressed so that it can be processed the next day. Because of the size of the object, it might be better not to decompress on first touch but to do the decompressing on off hours. See "Compressing/Decompressing Objects in a CLP" above for how to submit the decompressing to batch.
Every time an object(s) are compressed, ISM will produce an archive report. It might be a good idea to keep a copy of this report. If you execute the PAEZIP/ARCHIVE OPTION(*COMP) command all objects that were compressed will appear on the same report. If you use *C on the Archive file, then there will be a report for each *C.
Review a Sample Archive Compression Report
If you have compressed IBM objects, it is recommend that you decompress all IBM objects before updating to a new release of OS/400. Look in the archive file by executing PAEZIP/ARCHIVE. This will show you what objects are in the file. You can then decompress those objects by either using the *D in the library of the object or using the PAEZIP/ARCHIVE OPTION(*DECOMP) OBJECT(LIBRARY/OBJECTNAME). It is better to submit this command to batch. If you have a lot of IBM objects, submit so that they are processed in the evening.
Link to description of PTFs needed for ISM and Java 1.2
Depending on which release of JAVA you have loaded to the AS/400 and the version of OS/400 installed. Security can be different. To give the public all authority to a zipped file you can add a command like this to a CLP or ADDJOBSCDE. I believe V5R1M0 will address all of these issues.
CHGAUT OBJ('/paezip/*') USER(*PUBLIC) DTAAUT(*RWX) OBJAUT(*ALL)
The folder/directory QDLS and objects cannot be used with PAEZIP command or the ISM system. Java does not support QDLS.
| Home
| PAEZIP | ISM
| PC/Accelerator
Home | Saving Time on
*NONSYS/*ALLUSR | Disaster
Recovery Software|