Hide Forgot
Description of problem: ccs_sync does not return a non-zero exit code if something goes wrong, even when -w parameter is given. This was discovered during testing for IPv6 support where ccs_sync was unable to connect to any of the nodes. It still returned 0. Similarly, it should possibly exit non-zero when invalid parameter is given. [root@z4 cluster]# ccs_sync -i -w Unable to connect to z46 Unable to connect to z26 [root@z4 cluster]# echo $? 0 [root@z4 cluster]# [root@z4 cluster]# ccs_sync --invalid-option ccs_sync: invalid option -- '-' Usage: ccs_sync [options] [<host0> [... <hostN>]] -f <cluster_conf> Path to the cluster.conf file to propagate. -c <NSS certificate DB> Path to your cacert.pem file. -i Increment the configuration version before propagating. -h Print this help dialog. -p <port number> Connect to ricci on the specified port. -w Exit with failure status if any warnings are issued. -v Print version information. [root@z4 cluster]# echo $? 0 [root@z4 cluster]# Version-Release number of selected component (if applicable): ricci-0.16.2-35.el6 How reproducible: Always Steps to Reproduce: 1. create a cluster (IPv6 cluster for build =< 35 is a good case) or use any other technique to make sure ccs_sync won't be able to connect 2. run ccs_sync -i 3. print exit code using 'echo $?' Actual results: Returns 0 on error. Expected results: Returns non-zero on error.
This bug came out of https://bugzilla.redhat.com/show_bug.cgi?id=697493
How to test: Set up a 2 node cluster with ricci running on both nodes. Run ccs_sync -i -w it should succeed with exit value 0 [root@amoco-02 ~]# ccs_sync -i -w [root@amoco-02 ~]# echo $? 0 Stop ricci on one of the nodes: [root@amoco-01 ~]# /etc/init.d/ricci stop Shutting down ricci: [ OK ] Try ccs_sync on the node still running: [root@amoco-02 ~]# ccs_sync -i -w Failed to connect to amoco-01.lab.msp.redhat.com: Connection refused. [root@amoco-02 ~]# echo $? 0 After fix, it should look like this: [root@amoco-02 ccs_sync]# ./ccs_sync -i -w Failed to connect to amoco-01.lab.msp.redhat.com: Connection refused. [root@amoco-02 ccs_sync]# echo $? 1
http://git.fedorahosted.org/git/?p=conga.git;a=commitdiff;h=a96e1ff61ea5095864ca1b07a5a2af111d7c68a2
[root@ask-04 tmp]# ls -l /etc/cluster/cluster.conf ls: cannot access /etc/cluster/cluster.conf: No such file or directory [root@ask-04 tmp]# ccs_sync -i Unable to parse /etc/cluster/cluster.conf: No such file or directory [root@ask-04 tmp]# echo $? 1 [root@ask-04 tmp]# /etc/init.d/ricci stop Shutting down ricci: [ OK ] [root@ask-04 tmp]# ccs_sync -i Failed to connect to ask-04: Connection refused. [root@ask-04 tmp]# echo $? 1
Technical note added. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. New Contents: Cause: Ricci isn't running or another error is ocurring when using ccs_sync with the '-w' option Consequence: ccs_sync doesn't return '1' on failure Fix: Updated ccs_sync to return '1' on failure Result: ccs_sync -w now returns '1' on failure.
Just a clarification, this bz covers the case where ccs_sync -i -w (or ccs_sync -i) doesn't return '1' when it can't connect to one of the nodes on ipv4, other return code issues are in bz 818335.
Technical note updated. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. Diffed Contents: @@ -1,4 +1,4 @@ -Cause: Ricci isn't running or another error is ocurring when using ccs_sync with the '-w' option +Cause: Ricci isn't running on one or more nodes when using ccs_sync with the '-w' option Consequence: ccs_sync doesn't return '1' on failure
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-2012-0898.html