Red Hat Bugzilla – Bug 1264195
Typo in ovirtnode/storage.py reread_partitions causes RHEV-H install to fail
Last modified: 2016-03-09 09:38:52 EST
Created attachment 1074577 [details] screenshot of traceback Description of problem: Typo in ovirtnode/storage.py causes RHEV-H install to fail randomly with traceback: File "/usr/lib/python2.6/site-packages/ovirtnode/storage.py", line 251, in reread_partitions "secs") % (timeout -i) TypeError: unsupported operand type(s) for %: 'NoneType' and 'int' ovirt-node-3.2.3/src/ovirtnode/storage.py: --- 243 if "dev/mapper" in drive: 244 _functions.system("service multipathd reload") 245 _functions.system("multipath -r &>/dev/null") 246 # wait for device to exit 247 i = 0 248 timeout = 15 249 while not os.path.exists(drive): 250 logger.error(drive + " is not available, waiting %s more " + 251 "secs") % (timeout - i) 252 i = i + i 253 time.sleep(1) 254 if i == timeout: 255 logger.error("Timed out waiting for: %s" % drive) 256 return False --- Version-Release number of selected component (if applicable): ovirt-node-3.2.3-20.el6 How reproducible: unknown Steps to Reproduce: 1. Start install of RHEV-H 6.7 Actual results: Sometimes install fails with traceback as seen in attached screenshot Expected results: Install finishes successfully Additional info: The line: 251 "secs") % (timeout - i) should be: 251 "secs" % (timeout - i)) It's likely that the line: 252 i = i + i should be: 252 i = i + 1 as well.
Created attachment 1074578 [details] simple patch
Thanks for the patch. I had to modify it slightly, because the format string needed another pair of braces.
I change the QA Contact to myself to follow this bug further. I tested clean automatic installation and TUI reinstallation with build rhev-hypervisor6-6.7-20150828.0.el6ev and build rhev-hypervisor6-6.7-20150911.0.el6ev on ibm-3650m4-02(did not use remotely media via imm, just use PXE) and dell-R910(pxe) servers with FC HBA 6 times. I did not encounter installation failed issue and did not see the traceback error during my testing. According to this code, I did the following, here can see the detail in screenshot, but the device is forced reload quickly, can not easy to trigger " while not os.path.exists" # systemctl multipathd reload; multipath -r; ll /dev/mapper/3600a0b80005ada7600004b9455a5ddb0
Created attachment 1085105 [details] screenshot.png
Hi Derrick, We can not reproduce this issue on QE's test environment. QE verified the patches modification is valid. Could you help to provide the RHEV-H 3.5.6 build to customer to verify this bug once the build is built ready by devel? QE will do sanity testing as well for this bug fix. Thanks Ying
According to comment 9, Checked the code patch verification pass. Did the sanity testing on this bug on FC machine PASS. Sanity coverage scenarios: 1. clean automatic installation PASS 2. TUI clean installation PASS 3. TUI reinstallation PASS # cat /etc/redhat-release Red Hat Enterprise Virtualization Hypervisor (Beta) release 7.2 (20151221.1.el7ev) # rpm -qa ovirt-node ovirt-node-3.6.0-0.24.20151209gitc0fa931.el7ev.noarch
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. https://rhn.redhat.com/errata/RHBA-2016-0378.html