Rocket Mainframe DevOps

I’ve just installed the OpenSSL tool and am running a BPXBATCH job to display the version information using the following command: openssl version It works, in that it displays this message: OpenSSL 1.0.2h 3 May 2016 however on the next line it also produces this message: WARNING: can’t open config file: /rsusr/rocket/ssl/openssl.cnf In my case the tool is installed into a directory called /usr/local/Rocket/ported/openssl I can’t work out why the version command is trying to open /rsusr/rocket/ssl (which doesn’t exist on my system) Any ideas what I have done wrong or haven’t done?

Status: DeclinedPopulate PDSTGTBL when processing S1 function, development driven stage like-other.When building JCL for CA JCLCheck validation of like other (JCL components) it would be helpful to be able to access a list of other JCL components in the same and participating packages.  The CA JCLCheck remote validation requires creation of a control statement [SSPROC dsn(member)],.  PDSTGTBL provides this information when staging from the package member list (S2, package drivenstage like-other) but during S1, development driven stage like-other, processing the table is not available and contains unrelated information.

We’re currently running PHP version 5.4.4 build 0003 on z/OS V2R2. When running a simple php -v or php -help command we receive S0CC abend (HFP-Exponent-Overflow Exception). This problem occurs in one of our Test LPARs but we’re not seeing the issue on another one of our test LPARs. Could this problem be related to a specific LE Option? What can we do to further troubleshoot this issue?

Implemented perl 5.16.2 as part of z/OS 2.2 migration. A critical perl user was not able to run existing code with 5.16.2 but it works with 5.8.7. Here’s our test case: 5.8.7: #!/usr/lpp/perl #####!/usr/bin/perl open( RESPONSE_BASE, “<:encoding(ISO8859-1)”, “/u/snomre/weekly_tdi/kmbtstorig.txt” ); open( RESPONSE_MOD, “>:encoding(ISO8859-1)”, “/tmp/kmbtst.txt” ); while (<RESPONSE_BASE>) { chomp; print( “before substitution $_\\n” ); s/( ¬# \\s*)HELLO=.*/$1HELLO=Melissa/; print( RESPONSE_MOD “$_\\n” ); } close(RESPONSE_BASE); close(RESPONSE_MOD); 5.16.2 #!/usr/bin/perl open( RESPONSE_BASE, “<:encoding(ISO8859-1)”, “/u/snomre/weekly_tdi/kmbtstorig.txt” ); open( RESPONSE_MOD, “>:encoding(ISO8859-1)”, “/tmp/kmbtst.txt” ); while (<RESPONSE_BASE>) { chomp; print( “before substitution $_\\n” ); s/( ¬# \\s*)HELLO=.*/$1HELLO=Melissa/; print( RESPONSE_MOD “$_\\n” ); } close(RESPONSE_BASE); close(RESPONSE_MOD); Expected response in 5.8.7: before substitution # HELLO=abc

I am sure that most ZMF customers can appreciate that it has been a reasonably busy week or so here in Serena ‘ZMF land’. We therefore thought it fitting to recap events and provide a couple of updates that all customers who already have moved, or who intend to move, to ZMF 8.1.1 need to know.       ZMF 8.1.1 Essential Maintenance Every ZMF customer reading this article should now be aware of the ‘Essential Maintenance’ alert that was issued in relation to DEF284711.  If not, please familiarise yourself with the following article (you will need to be logged into the Serena Support website to access this link):http://knowledgebase.serena.com/InfoCenter/index?page=content&id=S141693&token=a0dd0414590c455ca7e9ea082e8d90ca We cannot overstate the importance of applying this maintenance to your 8.1.1 software prior to upgrade of your ‘live’ ZMF systems.  The reality of this problem is that you are highly likely to lose data if you move ahead with Z

I’m testing php 5.4.4 to replace the ibm ported php 5.1.2 under IBM HTTP server 8.5.5.9. with the IBM ported php, i have a cgi called php, which is a shell script that then executes the php cgi program, with the php script name from the url. With php 5.1.2, I can run a php script that does the print phpinfo() and I get the proper html page back. When i change my php shell script to use the 5.4.4 php_cgi, I don’t get any output back. Instead I get a 404 error page in my browser, but neither the web or php error logs show any problem. Have you run into this issue? What do I need to do in order to get this working with the 5.4.4? Thanks, Peter

I just installed Python and the thread module is missing. Is this expected? Other modules import successfully. $ echo $PYTHONHOME /usr/lpp/ported/bin $ echo $PYTHONPATH /usr/lpp/ported/lib/python2.7 $ python -vvv # installing zipimport hook import zipimport # builtin # installed zipimport hook # trying /usr/lpp/ported/lib/python2.7/site.so # trying /usr/lpp/ported/lib/python2.7/sitemodule.so # trying /usr/lpp/ported/lib/python2.7/site.py # /usr/lpp/ported/lib/python2.7/site.pyc matches /usr/lpp/ported/lib/python2.7/site.py import site # precompiled from /usr/lpp/ported/lib/python2.7/site.pyc # trying /usr/lpp/ported/lib/python2.7/os.so # trying /usr/lpp/ported/lib/python2.7/osmodule.so # trying /usr/lpp/ported/lib/python2.7/os.py # /usr/lpp/ported/lib/python2.7/os.pyc matches /usr/lpp/ported/lib/python2.7/os.py import os # precompiled from /usr/lpp/ported/lib/python2.7/os.pyc import errno # builtin import posix # builtin # trying /usr/lpp/ported/lib/python2.7/posixpath.so # trying /usr

All z/OpenSource languages and tools have in common the existence of a file named README.ZOS containing instructions and example commands for installation of the tool. This file is located in : share/doc/<toolName>/<version>/ There are also instructions for unpacking.

Peter Webb walks us through a mass rebuild of the package components in Changeman ZMF.#Videos#ChangeManZMF#Announcement

Hi, I’ve been working with the z/OS Open Source version of PHP and found a behavior that is different then the previous version we’ve used. We commonly use PHP scripts to parse through the results of system commands at our shop for system checking and monitoring. One such example is a script that checks the available space in file systems. We’ll do something like this: $dfcmd = “df -kP" exec(”$dfcmd",$dfout,$dfrc); foreach ($dfout as $df) { #Check each filesystem } In the previous version of PHP the results from the command would result in $dfout being populated as an array with each output line as a separate array element. This allowed the following foreach loop to parse through the results. In the new version, $dfout is instead populated with a long string with the EOL characters included. Is this the desired behavior? Thanks!

The version of a distribution is often not enough for uniquely identifying it for purposes of knowing whether or not you have the most up-to-date code. Sometimes, for the same tool version, Rocket will make available for download multiple iterations each with a different build number. When requesting support for a language or tool from Rocket be sure to include the name, version, and build number. The build number is part of the original distribution file you downloaded from Rocket. Also, in the unpacked distribution you will find a VERSION.ZOS file that contains this information within this sub-folder: share/doc/<toolName>/<version> (e.g. share/doc/bash/4.3) It is best to keep this file so you can uniquely identify your tool version.

Can PHP track sessions and timeout user after x amount of idle time in an Apache and WebSphere Application Server on a Z system environment.

As we approach the release date for ChangeMan ZMF 8.1.1,  here is a screen print of a complete list of the HLLX Exits available in this new release..    We would like to solicit input from the user base as to what HLLX exits should be considered next for inclusion in the product.  One area that we feel should be looked at is ERO and we would like to get feedback on that.  #ChangeManZMF#Announcement

I'm trying to create a Java App to approve a CMN ZMF Package without requiring users to logon to the mainframe and into CMN ZMF etc.... I was able to get the 'Sample' Webservices Java app (provided via the support site) to work. However, when I attempt to code the approval process/service (i.e. PackageLifeCycle.ApprovePackageRequest) I'm getting some odd errors in the Java code (i.e. fields don't appear to be defined correctly) and I can't even get to a test phase. Is there anyone out there that has successfully done this. If so, would you be willing to provide a sample similar to the 'Package Information' Sample provided by Serena? Thanks.#ChangeManZMF

Title says it all - Do we have a ChangeMan equivalent for Personal Computers or to bw used on Hercules? Which can be used "at home", for the home grown mainframes under GPL.Thanks,Join us: www.ibm-mainframes.com#ChangeManZMF

Status: DeliveredIncluded in ZMF 8.3Since users may use different package create methods (long, short....) and some of these skip panels, update the first package create HLLX exist(s) to be able to set values for panels that may not be displayed and thus skipping HLLX processesing that may be required.Since users may use different package create methods (long, short....) and some of these skip panels update the first package create HLLX exist to be able to set values for panels that may not be displayed and thus skipping HLLX processesing that may be required. Example for simple package we use MANUAL for install and require that INSTALL date be 3 moonths from current date of package create to give time to code, test before baseline without passing install date.  Depending on method for package create other than LONG HLLX can not set standards required. I would like to see if the first HLLX exit(s) used for package create can set the values for other panels that may not be processed but

Welcome to the Rocket z/OS Open Source Languages and Tools forum. This is a place where you can ask questions about our open source languages and tools and get answers from other mainframe professionals like yourself, as well as members of the Rocket team.

Has anyone experienced a higher than normal amount of bugs in ZMF V713C? We have had to implement no less than 5 vendor fixes or workarounds to resolve issues.#ChangeManZMF

Status: Already OfferedThis function is already available via the component browse apiAdd a new REST method to the ZMF Package Services that enables the browsing of the actual content of a package component to the REST services. In the old TT4ZMF integration, this functionality was available. It was used so that peers or dev leads could do peer code review of packages/components and then approve all from SBM. Functionality was lost when moving to ZMF Package Services. 

Status: DeclinedDeclinedWe need to scratch several components of the same libtype.  We would like to use the XML SCRATCH COMPONENT SERVICE.  It seems to only allow one member at a time.

The rapidly expanding growth of distributed ZMF clients can be attributed to a more intuitive user experience,  smart editors, integrated debuggers and more.   As this technology becomes more widely implemented,  customers have requested  stronger network security for their ZMF Environment.  Specifically the request has come in for all ZMF communications to be encrypted using the strongest version of TLS.   The Secure Sockets Layer (SSL) protocol was designed by Netscape in 1995.  SSL went through many iterations and has now been effectively replaced by Transport Layer Security (TLS).   The terms SSL and TLS have become somewhat interchangeable, but they both exist to provide secure transport for your communications.  We have all used TLS Encryption technology when accessing sensitive data over the internet by using the https protocol .    Implementing TLS for ZMF distributed clients is fairly straightforward. 

Has anyone modified exit 9 (CMNEX009) to exclude adding approver by application? Currently an approver is added whenever libtype equals XYZ. Working as designed. We would like to have exit 9 as I mentioned above also to exclude adding XYZ if application equals ABDC. Any assistance would be greatly appreciated.#ChangeManZMF

As much as we may hate to admit it, problems do sometimes arise in the ZMF Server started task that may lead to an abend (and I hasten to stress that these are not always Serena's fault!).  Also, back in the dim and distant past that coincided with the release of ZMF 7.1, the recommended started task dump file contained in the SERVER sample CNTL member changed from the tried and trusted: //SYSUDUMP DD SYSOUT=* *Abend list to the new (and slightly mysterious for some): //SYSMDUMP DD DISP=(MOD,CATLG,CATLG), * SYSMDUMP // DSN=somnode.SERCOMC.SYSMDUMP( 1), // UNIT=SYSDA,SPACE=(CYL,(200,100),RLSE), // DCB=(DSORG=PS,RECFM=FBS,LRECL=4160,BLKSIZE=4160) Many of our customers are comfortable with SYSMDUMPs.  However, if you are unfamiliar with this format, have never made the change to your started tasks’ JCL, are wondering why this happened, how

I am trying to trace why a promote job is not building a refresh of a WLM environment for a DB2 stored procedure. Is there something that can be added to a job that shows you variable values that impact what is and isn't included in JCL? I've been blessed with administration authority with no formal training. Thanks!#ChangeManZMF

If you have not installed ZMF 8.1 yet and tried the high level exits out you need to soon. In addition to giving us the flexibility to use COBOL, REXX or any other high level language, the developer have made it easy to get variables and even easier to pass variables between stages or even between sessions. Today I just want to share how to use REXX to read a file for validating the Work Request ID. -- It sounds simple but until now unless you were able to program in assembler or only accessed ZMF through the ISPF interface it wasn't possible. The sample code is just that -- probably not that useful in a production environment -- but it depends. Remember that a REXX program can't read a VSAM file so you have copy it to a flat file or use another language. Also since there is no TSO environment in the high level exit you have to use BPX WDYN to allocate and free files - this is a standard utility on all mainframes. Here is the example of that call: RC = BPXWDYN("ALLOC DD("ddname") DA('