Red Hat Bugzilla – Bug 1009897
Adding zFCP device fails on s390x guest with one CPU
Last modified: 2014-10-14 03:35:27 EDT
Description of problem: On a z/VM guest with one CPU, anaconda fails to add a zFCP LUN for the first time user tries it. Following attempt using the same device_number/WWPN/LUN is successful. Version-Release number of selected component (if applicable): anaconda-13.21.195-1.el6 (RHEL-6.4 GA) anaconda-13.21.206-1.el6 How reproducible: * always on a guest with one CPU and no DASDs (6 attempts) * sometimes on a guest with one CPU and 3 DASDs * seen twice on a system with two CPUs, hard to reproduce Steps to Reproduce: 1. start vnc installation on an guest with one CPU 2. proceed to Specialized Storage Devices screen 3. go to Add Advanced Target - Add ZFCP LUN 4. enter device number, WWPN and LUN and click Add 5. see error message: zFCP device 0.0.a007 not found, not even in device ignore list. 6. click OK and try to add the same device again - it will succeed. Actual results: zFCP device is not added, user has to repeat adding of the device Expected results: zFCP device is added successfully
Code from storage/zfcp.py: ... if not os.path.exists(online): log.info("Freeing zFCP device %s" % (self.devnum,)) iutil.execWithRedirect("zfcp_cio_free", ["-d", self.devnum], stdout="/dev/tty5", stderr="/dev/tty5") if not os.path.exists(online): raise ValueError, _( "zFCP device %s not found, not even in device ignore list." %(self.devnum,)) ... Reproduced manually: [anaconda root@rtt7 root]# zfcp_cio_free -d a107; ls /sys/bus/ccw/drivers/zfcp/; sleep 1; ls /sys/bus/ccw/drivers/zfcp/ ls: cannot access /sys/bus/ccw/drivers/zfcp/: No such file or directory 0.0.a107 bind module uevent unbind [anaconda root@rtt7 root]# I'm not sure if this is problem in anaconda/s390-utils/kernel, please reassign if necessary.
This request was not resolved in time for the current release. Red Hat invites you to ask your support representative to propose this request, if still desired, for consideration in the next release of Red Hat Enterprise Linux.
I don't believe anaconda is the correct component for this; the issue seems to be at a lower level than the installer. My assumption would be s390utils; reassigning.
Jan, I guess it's the same issue as in 1054691.
Yes, it looks so.
Not able to reproduce any more with s390utils-1.8.2-54.el6 on a system with one CPU: [root@rtt7 ~]# cat /proc/cpuinfo vendor_id : IBM/S390 # processors : 1 bogomips per cpu: 2913.00 features : esan3 zarch stfle msa ldisp eimm dfp etf3eh highgprs processor 0: version = FF, identification = 11C047, machine = 2827 [root@rtt7 ~]# [root@rtt7 ~]# rpm -q s390utils s390utils-1.8.2-54.el6.s390x [root@rtt7 ~]# [root@rtt7 ~]# zfcp_cio_free -d a107; ls /sys/bus/ccw/drivers/zfcp/; sleep 1; ls /sys/bus/ccw/drivers/zfcp/ 0.0.a107 bind module uevent unbind 0.0.a107 bind module uevent unbind [root@rtt7 ~]# Also adding zFCP disk during installation works fine. Not moving to VERIFIED in case you want to run additional tests.
Switching to VERIFIED as jstodola already verified this.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2014-1546.html