You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Apr 23, 2020. It is now read-only.
The script install-SMPE-PAX.sh is expecting AZWE001.readme.txt to be in ASCII form on zOS and convert it to EBCDIC as one of it's steps. However, the documentation on
tells user to "upload the AZWE001.readme.txt file in text format". When user upload it as text format, it will be changed from ASCII to EBCDIC causing the script to fail. The script should expect the readme file to be in EBCDIC on zOS and skip the convert to EBCDIC step to be consistent with user expectation.
@jackjia-ibm When the Jenkins pipeline copies text files from the build machine to River or Marist for testing, does it convert ASCII to EBCDIC? If not could that be done, so that the install-SMPE-PAX.sh script can have an EBCDIC file just like the real customer would?
The README has to be converted to EBCDIC somewhere. The customer is advised to do that at transfer time e.g. FTP it in text format, provided FTP recognises the EBCDIC code page setting of the target system. If you don't do it at this point, you can view it in ISPF as ASCII. Its main purpose is to guide the person performing a manual install. That person will normally be a sysprog, who will want to run all the jobs manually, so they won't use the script.
However, the install-SMPE-PAX.sh script reads the README automatically to extract the GIMUNZIP job. It assumes the README is in ASCII so it converts it - otherwise you need a convert step before the script, or need to have transferred it as text converted to EBCDIC. That's why I put the convert step inside the script.
The Jenkins pipeline also runs the install-SMPE-PAX.sh script, so it does not convert the README to EBCDIC when it uploads it.
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
@kathysychan commented on Wed Sep 25 2019
The script install-SMPE-PAX.sh is expecting AZWE001.readme.txt to be in ASCII form on zOS and convert it to EBCDIC as one of it's steps. However, the documentation on
https://docs.zowe.org/stable/user-guide/install-zowe-smpe.html#upload-the-download-package-to-the-host
tells user to "upload the AZWE001.readme.txt file in text format". When user upload it as text format, it will be changed from ASCII to EBCDIC causing the script to fail. The script should expect the readme file to be in EBCDIC on zOS and skip the convert to EBCDIC step to be consistent with user expectation.
@kathysychan commented on Wed Sep 25 2019
@John-A-Davies for your awareness
@John-A-Davies commented on Wed Sep 25 2019
@jackjia-ibm When the Jenkins pipeline copies text files from the build machine to River or Marist for testing, does it convert ASCII to EBCDIC? If not could that be done, so that the
install-SMPE-PAX.sh
script can have an EBCDIC file just like the real customer would?@jackjia-ibm commented on Wed Sep 25 2019
@John-A-Davies yes the iconv step can put in
temp-fixes-before-install.sh
script to properly prepare for the install.The text was updated successfully, but these errors were encountered: