New way in RPG to monitor for errors

In RPG we have been able to monitor for errors, using a Monitor group, for years. Until IBM i 7.5 and 7.4 TR6 we could only monitor for program and file status codes. I have always preferred CL’s MONMSG that allowed me to monitor by message id instead.

As part of the new release, or new TR for 7.4, there is a new RPG operation code, ON-EXCP, that can be used within a Monitor group to monitor for a specific error message or messages. The syntax is very similar to that of the ON-ERROR operation code:

monitor ; Something on-excp 'CPF1234' : 'RNX1234' : 'RNQ1234' ; Do something else to mitigate the error endmon ;

I am going to give some examples below so you can understand how ON-EXCP works. Here is my RPG program’s source code:

New TR for IBM i 7.3

I am certain in all the videos of the announcements for IBM i 7.5 and 7.4 TR6 we were told there would be no more Technology Refreshes, TR, for 7.3.

Well, it appears that there is to be another Technology Refresh for 7.3, TR12. I found it on the Technology Refresh information page for 7.3 here.

It is available on the same day, May 24, that 7.4 TR6 is.

What is available in this TR for 7.3?

Making a list of journal receivers with the detach date

Update

A SQL View was added to IBM i 7.5 that makes it easier to retireve information about Journal Receiver. Before using what is described below check my post about getting Journal Receiver information with SQL.

 

---

 

I wanted to write a program that would delete old journal receivers, ones detached from the journal and more than a certain number of days old (to be determined later). I can see the detach date when I use the Display Journal Receiver Attributes command, DSPJRNRCVA, but I can only do this for one receiver at a time, and the command only allows for display or print. Looking at my chart of SQL Views and Table functions I found that is not one for Journal Receivers. Alas, this leaves me having to use an API to get to this information.

The information I desire is:

• Journal name and library
• Journal receiver name and library
• Attach date (timestamp) of the receiver
• Detach date (timestamp) for the receiver

There is an API that will return to me the information. It comes with the long name: QjoRtvJrnReceiverInformation. Before I go into the example of using this API let me get started with how I made a list of all the receivers in a library. Here I can use a SQL Table function, OBJECT_STATISTICS. I use the following statement to display a list of all the receivers in a library.

SELECT OBJNAME,OBJLIB FROM TABLE(QSYS2.OBJECT_STATISTICS(<library name>,'*JRNRCV','*ALLSIMPLE'))

IBM i Merlin demo

Demonstration of IBM i Merlin that was announced yesterday as part of IBM i 7.5 and the latest Technology Refresh for 7.4, TR6.

New version 7.5 announced

Most of us knew it was coming, so there is no surprise that a new release for IBM i has been announced. The new release, 7.5, accompanies a Technology Refresh update for version 7.4 TR6. As only two versions of the IBM i are supported with TRs it does also mean the end of TRs for version 7.3. If you are using a partition with 7.3, and you can upgrade, this should be the time to consider doing so to at least 7.4.

IBM i 7.5 will only run on Power 9 and 10 servers.

Many of the new features of 7.5 will not be supported by 7.4 TR6. I will make note here of which are additions or changes that apply to 7.5 only.

Placing SQL functions into the SQL Insert statement

To write the posts for this website I need to have examples DDL tables. I often use a Table whose first column contains the row number and a second contains some kind of random number.

While I was illustrating how to fill one of these tables it struck me that there was an easier way to do this, having all the logic in the Insert statement itself.

This is the source code for the Table I commonly use:

01 CREATE TABLE MYLIB.FIRST_TABLE 02 FOR SYSTEM NAME "TABLE1" 03 (FIRST_COLUMN FOR COLUMN "FIRST" VARCHAR(10), 04 SECOND_COLUMN FOR COLUMN "SECOND" SMALLINT, 05 UNIQUE (FIRST_COLUMN)) ;

Printing information about SQL RPG programs

I am using the word "program" in a generic way here. In reality what I am writing in this post can also be used for SQL packages (object type *SQLPKG) and service programs too.

What I wanted to know was which SQL statements executed within a program, and the tables and indexes these statements used. Fortunately, there is a command for that: PRTSQLINF, Print SQL Information. Before I show what this command can do, I need to have a table and program I can use to create the information that the command will return.

First I need a table. I know I will not win awards for originality of this table's name; I have called it TESTTABLE.

01 CREATE TABLE MYLIB.TESTTABLE 02 (FIRST_COLUMN FOR COLUMN "COL1" CHAR(10), 03 SECOND_COLUMN FOR COLUMN "COL2" INT) ;

ACS 1.1.9.0 is replaced

 

The original contents of this page have become obsolete, go to this page for up-to-date information.

 

List rows in IFS files that contain a certain string

The question was an interesting one: How is it possible, in SQL, to list all of the files in the IFS that contain a certain string?

There are a number of SQL table functions that allow me to get to information about the files in IFS folders, and others to "read" their contents. This is a good excuse to combine the two.

I am going to place three files in my personal IFS folder, MyFolder, then search those for the desired string. I created three DDL tables in my library, MYLIB. OK, I created one table and then cloned it two times to make the three tables. I populated them with just a few rows of data. As the tables are identical, I can easily combine the results from the three in one SQL statement using the UNION clause:

SELECT 'TABLE1',A.* FROM TABLE1 A UNION SELECT 'TABLE2',B.* FROM TABLE2 B UNION SELECT 'TABLE3',C.* FROM TABLE3 C ORDER BY 1,2

World backup day

A day like today is a good reminder of the importance of backups. I am sure we have all had to restore data from a backup many times in our lives in IT.

Do you know if your company's backup are valid? It always surprises me how many company's purchase a new device for backups, tape drive or VTL, and never try to do a test restore. Trust me you want to make sure you know how to restore data from the device before your really need to.

Not really part of backups but make sure you only retain backups for the time they might be needed, this varies by industry and would imagine by country too. Once the backup has passed the retention date destroy it. I have been through an e-discovery process, it became an almost full-time job for a month to find every backup that was on premise, off site, and even in the IT members cars and homes. It gets expensive too when you are asked to restore data from tapes you no longer have a tape drive for, and you have to contract with a third party to do it on your behalf. It is so much easier to tell the auditors that, in agreement with your company's legal department and the written company policy, all backups are destroyed after the designated period of time.

SQL indexes suggested from the job log

I have written before about how Db2 for i will suggest indexes, using the SYSIXADV view or Navigator for i, that will improve the performance of your SQL programs, etc. There are times I need to know if any indexes are advised when a program has run just once.

Fortunately this information is written to the job log, when the statement is executed in the program.

Before I demonstrate that I need a SQL DDL table I can use:

CREATE OR REPLACE TABLE MYLIB.TABLE1 (FIRST VARCHAR(20), SECOND VARCHAR(20)) ON REPLACE DELETE ROWS

Notice that the table does not have a key.

Changes to the RFE process

This morning I received an email from IBM informing that there are changes coming to the RFE, Request For Enhancement, process.

IBM uses the RFE process to drive a lot of the enhancements they add to new releases and Technology Refreshes to all of their products, including IBM i. If you have an idea you think would be a valid improvement to IBM i you would a submit an RFE. Other people can vote on the RFE if they agree it would be a worthwhile enhancement.

I am not going to repeat what IBM has already published on their Community web site. What I am going to do is to give you this link to it here.

If you have a bookmark in your browser to the RFE old web page you will need to delete your existing one and create a new one as it is now found at a different URL.

I encourage you all to submit RFEs if you have any ideas, and vote on others. If you don't IBM will never know that wonderful idea you just came up with.