Saturday, July 22, 2017

Backup your AS400 SAVF in your PC

Hello Friends,

Taking backup is one of the most important thing in project life cycle. When it comes to AS400 we do save our program & data objects into TAPE using SAVOBJBRM commands. But we can also take a copy of our works (program & data objects) and save it in SAVF. And this SAVF can be transferred into your desktop.

This could be useful if you want to transfer any objects from one AS400 to another AS400 but both are not talking to each other. (If both servers are in same network then we can just copy the data using any ways like DDMF or SNDNETF or FTP)

Downloading your SAVF into your PC:

Step 1: Save your objects into SAVF

We are creating a SAVF called BKUP and saving object TESTDATA inside the SAVF

Now before we look on how to transfer this to your desktop, we need understand the IFS representation of an object in AS400.

Now our SAVF object DEVYUS1/BKUP can be viewed/identified as IFS path using below command


In similar fashion we can see all our PF, Programs and other objects. This is important because using this we are going to FTP our BKUP.FILE into our desktop.

Step 2: Download BKUP SAVF into our desktop

When we login the current IFS directory is our home directory (/Home/DEVYUS) then we change that using Change Directory (CD) command as /QSYS.LIB/DEVYUS1.LIB

Then we change our desktop local directory (whatever you wish)

Then we choose BIN (Binary) mode to receive the file (SAVF can be transferred using BIN mode)

Then just GET command will receive the file and store it in your local directory

How to upload the file again to AS400:

Again it is quite simple and instead of GET we need to use PUT command to upload the file. But before that we need to create the SAVF in AS400 server manually.

Let us try to upload this SAVF in another library DEVYUS2

Step 1: Create SAVF


Step 2: Upload using FTP

First we change our Current Directory of AS400 to “/QSYS.LIB/DEVYUS2.LIB”

Then we change our local directory of our desktop

Then changing the transfer mode to BIN (Binary)

Then using PUT command we transfer file back to AS400

So that’s it, hope you got the idea of how to send/receive SAVF from/to your IBM I server.

Have Fun…!!! Happy Coding…!!!

Friday, July 21, 2017

Common Compilation Errors in RPGLE

Hello Friends,

Today we are going to see some of the common errors we come across while compiling the RPGLE code. If you are beginner then this post could be useful if you come across these error messages in your future.

It always give happiness when you compile you code with error code 0 on first time J

Sample code with Error:

Order file:


Basically this program gets Ord_No & Cust_No as input and checks if the order is available in ORDER file. If available then it prints the Order quantity else it writes new record for the customer and order number with other fields (Name & quantity) as default.

After compiling this below is the list of errors:

Let us go one by one and fix it.

RNF7030 à ORD_QTY is not defined

The intention of this line is to display order quantity from file but variable used is Ord_Qty but file contains field as ORDQTY. So compiler is not able to recognize Ord_Qty. Thus we need to remove this underscore

Eval      Msg = 'Order is available '+
                'Qty : ' + Ord_Qty    

RNF7055 à Factor 1 ORD_KEY is not valid for the specified operation.

Ord_Key       Chain     Order

Here we are trying to do chain operation on the file Order but the file specification is not having record access as Key (‘K’)

FOrder     IF   E             Disk  à This should be changed as below
FOrder     IF   E           K Disk

RNF7016 à Display length 100 greater than maximum allowed of 52;

This is because we are trying to display Msg variable using DSPLY function. But the Max length that could be displayed is 52. But Msg variable is having length 100. We need to change this to Max 52.

RNF7416 à The types of the right and left hand side do not match in EVAL operation.

This is most common error you might have already seen it.

Eval      OrdName  = Ord_No     // we are trying to eval char variable with numeric

RNF5196 à File in Factor 2 does not have allowed file type for WRITE

When we need to add new records, the file specification should be defined with File Addition = ‘A’

FOrder     IF A E           K Disk 

Ok now let us quickly make the changes and compile again. 

Now we eliminated all previous errors and ended with new 2 errors. Let us fix these.

RNF7073 à KFLD at sequence number 18 is 4 long. Key field is 5 long.

This is because, our key Ord_Key is formed using Ord_No & Cust_No variables with 4 and 10 length. But the actual files key is ORDNO (5 length) and CUSNO (10 length).

So to fix this, either we can change length of Ord_No from 4 to 5, else we can use Like keyword during 
declaration to ask program to refer data type directly from PF.

RNF7421 à Operands are not compatible with the type of operator.

This is because we are trying to concatenate string and OrdQty in Msg variable. But OrdQty is numeric and thus we need to convert it to %Char before concatenate

So the final code is here.

These are very basic errors and there are more to come when you use Copy book and Procedures and introduce other ILE concepts.

I will try to cover few more in my future posts until then

Have fun..!!! Happy coding..!!!

Saturday, June 10, 2017

Security problems with RMTCMD

Hello Friends,

This is a continuation of my earlier post on “how to execute CL command from Windows”.

If you have read that post you should probably know how RMTCMD works with AS400. But still there are chances of few security issues/violation using this method.

If I want to say in one liner, “using RMTCMD a user can run CL command directly into AS400 even if the user profile has “Limit Capabilities = *YES” LMTCPB(*YES)”

What is limit capabilities?

If a user profile is created with limit capabilities as yes, then that profile will not be able to use most of the CL command in the command line. When you try to execute a command you will get error as

“Command CALL in library *LIBL not allowed”

Using this we can limit our user group from accessing unwanted/sensitive commands.

E.g.: If someone issues a command DLTF on a PF this would probably delete the file from production.

But when we use RMTCMD option to execute this command from Windows

Can we control which command will be allowed for Limit capable users?

Yes, we can. Every CL command will be having attribute “Allow limited user”.

If it is “*No” then we cannot use this command using limited access profile

If “*Yes” then we can use this even if the profile has limited access

Note: But as a system administrator, you can change any CL command that can be run at command line to allow limited users to also run. This can be done using CHGCMD option.


But the rmtcmd.exe completely ignores the Limit Capabilities attribute of the user profile and therefore allows any user to run any command he or she is authorized to run. 

How to control this?

We have to use user exit programs in exit points to control any unauthorized entries.

Exit Programs:

An exit point is simply a point in an application at which the application can optionally call an external program to perform customized processing. In other words, whenever the RMTCMD is executed we can get the login details and validate it before it gets processed. For this we need to add our exit program into IBM exit points interface of RMTCMD.

Once you have registered your exit program, whenever a user attempts to log on using RMTCMD, the server finds your program that's registered for the exit point, then calls your exit program, passing as parameters information about the user who's logging on. Your exit program then processes that information and takes the appropriate action, according to the security rules you implement in the exit program. Upon return, your exit program passes back a flag to either ACCEPT or REJECT the logon attempt.

Exit Points:

Each exit point has a name and an Exit Point Interface. The Exit Point Interface is a list of input and output parameters the IBM server program exchanges with your exit program. The QIBM_QZRC_RMT exit point occurs immediately after a user enters a user ID and, password to log on to the AS400 server using RMTCMD command.

In this post I just wanted to call out the security issues with respect to RMTCMD. I would try to catch up with some practical example on how to use Exit points & Exit programs later.

Until then… Have Fun..!!! Happy Coding..!!!

Friday, June 9, 2017

Executing AS400 CL Command from Windows

Hello Friends,

If you have IBM System I Access for Windows then it would have given you a DOS command called ‘RMTCMD’. Using this you can directly run the CL commands from your Windows command prompt.


RMTCMD supports two formats


RMTCMD /I <filename> [/Q] //AS400SERVER [/Z]

RMTCMD à command name (CL commands can be given in double quotes)
AS400SERVER will be your server name

/Q à It specifies that command will display the error and end the command (if any error encountered in CL command). If we didn’t give then it will ask for options from you to continue or end

/Z à Specifies to display only required messages on your workstation

E.g.: I want to clear a file DEVYUS1/TEMP

It will prompt you for user name & password. And the commands will execute directly on your AS400.

E.g.: Output 

We can see using /Z option the extra details (version and copyright info) were not displayed.

What if we have multiple commands to execute? 

We can have it in a file and use our 2nd command.

What if we want to execute a DOS command from AS400?


You have to start PC Organizer (STRPCO) first then using STRPCCMD we can execute any DOS command.

Hope you might have learnt something useful. I shall come back with another interesting topic later. Until then… Have Fun..!!! Happy Coding…!!!

Sunday, May 28, 2017

Pointers in AS400 (IBM i)

Hello Friends,

Pointers are one of the key features in AS400 RPG programming. If you are a beginner and just started to write some RPG code then it is good to understand the concept of pointers.

What is pointer?

Pointers are used to hold the memory address of a variable (just like how it used in C). This will become handy when we need to process larger size data.

Let us take a simple example:

Pointer can be declared using asterisk (*) symbol. We can assign this pointer to a variable using Based keyword. In the below example P_Data is based on pointer Ptr. This means, Ptr will hold any address and P_Data will display the content of that address.

We can see from above that, address of Var1 is assigned to Ptr. Now Var1 & P_Data points to same address. So if we change Var1 data, it will be reflected in P_Data.

Similarly, if we change P_Data value then it will be reflected in Var1 value.

Segmentation with pointers:

Using pointers we can split the large data into small chunks and access it.

Let us take below example:

Here Var1 is of length 100 but we are accessing the data by splitting into small chunks (of 10 length) using P_Data variable.

We initially assign address of Var1 into Ptr and then we increment the Ptr value by adding size of P_Data.


Do you know this trick?

Hopefully you might have used RPG calling another RPG using parameter. We would have noticed same variable used for both input & output. (i.e. if PGMA calls PGMB has parameter ABC, then I change the value of ABC in PGMB then it will reflect in PGMA automatically).

This is because; the two-way parameter is achieved using "shared memory".

When one program calls another, the only thing that’s passed between them is an address in the computer’s memory where the parameter starts. Nothing else is passed.

Look at the below example for better understanding.

TESTPGM à calls GETNAME it passes Name as parameter. But length of Name is 10 in TESTPGM. But we receive this value using *Entry Plist under FullName variable (of length 20) in GETNAME program. We assigned value “Mohammed Yusuf” and program returned the value to calling program (TESTPGM).

Now, if we display variable Name we get “Mohammed Y” alone. The remaining value gets assigned to Address variable (since both are under DS)


Name = “Mohammed Y”
Address = “usuf”

What happened here? When program “GETNAME” called, system passed the address of Name variable. And GETNAME program assigned the value “Mohammed Yusuf” in that address.

Now when TESTPGM receives it, Name variable can hold only 10 length so remaining text is present in Address variable. (because Name & Address are in single Data Structure so it forms continues memory allocation)

How to overcome this?

Instead of *Entry Plist, start using PI (Procedure Interface)
  1. It is similar to *Entry Plist (but better than that)
  2. It requires matching prototype to work
  3. Replacement of *Entry Plist in Free format

The program GENAME is changed to GETNAMEPR and uses PI to receive the parameter. The calling program TEST is changed with Prototype definition of GETNAMEPR.

With this in hand, if we try to call GETNAMEPR with variable Name it will throw error since length of Name is not matching with parameter length of Prototype.

Parameters & Prototype has many useful things in ILE programming. I will try to cover those in my upcoming posts.

Until then, Have fun…!!! Happy coding…!!!

Monday, May 8, 2017

XML parsing in AS400 (IBM-i) using XML-INTO

Hello Friends,

If you are new in working with XML in AS400 or just starting to do some XML parsing, this article will be useful to you.

XML-INTO is the op-code used to parse the XML document and retrieve the values directly into the variables in RPGLE.

Previously I saw many examples using DOM parser but it is complex for any beginners. But IBM introduced XML-INTO op-code and made the XML parsing much easier. Let us directly jump into syntax and examples.


xml-into <VariableName> %xml(<filename> : options)

There are couple of options which are considered while parsing the xml. We will see one by one.

Mostly the <variablename> will be a data structure which matches with the xml structure. 

Simple XML:



Options in xml-into:


In general XML tag names should match with RPG variables with case sensitive. But using case=any option we can eliminate it.


This tells the op-code that xml data source is coming from a file document. In this case, the first parameter of %XML function will be file name.


Sometimes if we are having missing fields in XML document


Sometimes if we get extra tag in the XML (not defined in DS)

In other words, if u expect all the DS fields are mandatory in your XML you can give allowmissing=no and handle the exception and allowextra=no will not accept any extra tag to come in your xml.

In most cases, we use both as ‘yes’ and handle the exception in program.


Using path option you can directly read any particular tag value. We will see it deeper in coming examples.

A complex XML:

Let us look at below example which has nested tags, attribute value and also repeating a particular tag number of times. 

Here we need to know how to define a nested data structure and how to find number of times a tag being repeated. Fortunately, xml-into offers an easy way to do that.

And trust me we can just parse all the values with single statement J

How to form nested Data Structure:

If you look closely, we have split the group of tags into separate DS and nested into main Customer DS using LikeDS keyword. The reason for CtInfo_T & Item_T is declared as Template but Order_T & OrdDet_T are declared as Qualified because, if we are going to have any nested DS then the parent DS should be declared as Qualified.

Take a note on DepartmentNumber. This exceeds normal variable length of 15 char while defining in RPGLE. So if we have variable length more than 15 char then we have to write the full name followed by three dots. Thus system will look for its declaration in next line. (in our case 2S 0)


Do notice this variable Count_Item is an extra field apart from XML fields. When we need to parse multiple occurrence tags (in our case <item> tag) then we have to define a variable next to repeated tag so that system will automatically give us the number of repetition in this variable.

RPGLE code:

Please note on the countprefix option. We are saying the system to return the count of repetition under the variable has prefix “count_”

i.e. for item tag the count variable should be “count_item”

if we change “countprefix=cnt_” then in our DS, the variable should be “cnt_item”

here we go, we got all the values, and also the count_item has number of repetition as “3”.

With this in hand, we can very well do our program logic and proceed further.

One last example using path option:

Using path we can directly read any segment of xml. For example, I want to retrieve OrderNumber alone from the XML. Then,

Do note that if we are fetching value of a single tag directly then variable name to receive the value may not need to be the same xml tag name. (Here OrdNum is used).

Hope this post gives you enough idea to get start with xml-into method of parsing. It is really very useful as you can see it requires very minimal amount of coding. I will come with another topic soon. Until then…

Have fun..!!! Happy coding…!!!