Hide Forgot
2 What is the nature and description of the request? - Change nfslock init script in order to provide an error code and/or message if portmap is not running - If portmap is not started before nfslock, service is started without reporting any error but it's not really started. [root@dhcp-1-117 ~]# service nfslock restart ; echo $? Stopping NFS statd: [FAILED] Starting NFS statd: [ OK ] 0 3 Why does the customer need this? - To improve the user experience and in order to monitoring init scripts 4 How would the customer like to achieve this? - Reporting an error code/message once nfslock is started and portmap is not running 5 For each functional requirement listed in question 4, specify how Red Hat and the customer can test to confirm the requirement is successfully implemented. - Checking that nfslock init script to report an error code/message 6 Is there already an existing RFE upstream or in Red Hat bugzilla? - https://bugzilla.redhat.com/show_bug.cgi?id=534133 is related in someway. 7 How quickly does this need resolved? (desired target release) - Next release 8 Does this request meet the RHEL Inclusion criteria - Yes 9 List the affected packages - nfslock 10 Would the customer be able to assist in testing this functionality if implemented? - Yes, if it's needed
When portmap is turned off: ------------------------------------------------------------ # cat /etc/redhat-release Red Hat Enterprise Linux Server release 5.6 (Tikanga) # rpc.statd ; echo $? 0 # tail -n 2 /var/log/messages Feb 18 00:54:04 kaedila-r5 rpc.statd[20261]: Version 1.0.9 Starting Feb 18 00:54:04 kaedila-r5 rpc.statd[20261]: unable to register (statd, 1, udp) # ps -ef | grep [r]pc.statd ------------------------------------------------------------ It's not the init script's fault that rpc.statd returns 0 even when it exits with an error.
(In reply to comment #2) > When portmap is turned off: > ------------------------------------------------------------ > # cat /etc/redhat-release > Red Hat Enterprise Linux Server release 5.6 (Tikanga) > > # rpc.statd ; echo $? > 0 > > # tail -n 2 /var/log/messages > Feb 18 00:54:04 kaedila-r5 rpc.statd[20261]: Version 1.0.9 Starting > Feb 18 00:54:04 kaedila-r5 rpc.statd[20261]: unable to register (statd, 1, udp) > > # ps -ef | grep [r]pc.statd > ------------------------------------------------------------ > > It's not the init script's fault that rpc.statd returns 0 even when it exits > with an error. I agree with this.. the rpc.statd is exiting with a status of 1 and the init script is just return what its return to it daemon rpc.statd "$STATDARG" RETVAL=$? reassigning to initscripts
Back to nfs-utils - according to comment #2, the issue is that rpc.statd *isn't* exiting with a status of 1 when portmap isn't running. It's exiting with a zero status.