Description of problem: ======================= used rhs-hadoop-install - setup_cluster.sh for hadoop set up without creating required user first. Scripts has created users but hose users do not have same uid always. Version-Release number of selected component (if applicable): ============================================================= gluster - 3.6.0.12-1.el6.x86_64 rhs-hadoop-install-1_18-1.el6rhs.noarch rhs-hadoop-2.3.2-2.noarch How reproducible: ================= 2/4 Steps to Reproduce: 1. (do not add required user before starting setup) from management server create 2x2 dist-rep volume using two server as below ./setup_cluster.sh --verbose rhs-gp-srv7.lab.eng.blr.redhat.com:/brick1:/dev/vg_new/thin_vol1 rhs-gp-srv4.lab.eng.blr.redhat.com:/brick1:/dev/vg_new/thin_vol1 rhs-gp-srv7.lab.eng.blr.redhat.com:/brick2:/dev/vg_new/thin_vol2 rhs-gp-srv4.lab.eng.blr.redhat.com:/brick2:/dev/vg_new/thin_vol2 2. it erroed out with uid mismatch error Actual results: found following output <snippet> iptables configured and saved group hadoop added with GID=492 1 new Hadoop groups added user mapred added with UID=495 user yarn added with UID=494 user hbase added with UID=493 user hcat added with UID=492 user hive added with UID=491 5 new Hadoop users added ... ... group hadoop added with GID=497 1 new Hadoop groups added user mapred added with UID=498 user yarn added with UID=497 user hbase added with UID=496 user hcat added with UID=495 user hive added with UID=494 </snippet> log has following message:- [22:27:07] DEBUG: check_uids: "mapred" user has inconsistent UID across supplied nodes. UIDs: 495 498 rhs-gp-srv4.lab.eng.blr.redhat.com has mapred UID: 495 rhs-gp-srv7.lab.eng.blr.redhat.com has mapred UID: 495 rhs-client18.lab.eng.blr.redhat.com has mapred UID: 498 "yarn" user has inconsistent UID across supplied nodes. UIDs: 494 497 rhs-gp-srv4.lab.eng.blr.redhat.com has yarn UID: 494 rhs-gp-srv7.lab.eng.blr.redhat.com has yarn UID: 494 rhs-client18.lab.eng.blr.redhat.com has yarn UID: 497 "hbase" user has inconsistent UID across supplied nodes. UIDs: 493 496 rhs-gp-srv4.lab.eng.blr.redhat.com has hbase UID: 493 rhs-gp-srv7.lab.eng.blr.redhat.com has hbase UID: 493 rhs-client18.lab.eng.blr.redhat.com has hbase UID: 496 "hcat" user has inconsistent UID across supplied nodes. UIDs: 492 495 rhs-gp-srv4.lab.eng.blr.redhat.com has hcat UID: 492 rhs-gp-srv7.lab.eng.blr.redhat.com has hcat UID: 492 rhs-client18.lab.eng.blr.redhat.com has hcat UID: 495 "hive" user has inconsistent UID across supplied nodes. Expected results: User should have same gid and uid across cluster Additional info:
Code in setup_cluster.sh related to creating users (via useradd and via ldap/ipa) has been removed.
Checking rhs-hadoop-install-1_32-1.el6rhs.noarch Scripts automating non-ldap user setup: /usr/share/rhs-hadoop-install/bin/add_users.sh /usr/share/rhs-hadoop-install/bin/add_groups.sh are still shipped in the package. This contradicts the statement in comment 3 and our current documentation. Note: this is the same kind of problem as with BZ 1105149: it's not a good idea to ship unused code to the customers.
I forgot about add_users.sh and add_groups.sh -- good find! I've moved these to scripts to the archive github folder. I want to preserve the "user adding/ldap-ipa" code in setup_cluster. What is recommend practice here? I've commented out the call to the related functions and added a comment in these function that they are not used in Denali. But, I'd still like the code to be available if we decide, post-Denali, to automate some aspects of Hadoop user creation.
I have deleted all code related to creating the hadoop users. Fixed in version 1.34.
Tested with: rhs-hadoop-install-1_34-1.el6rhs.src.rpm (https://brewweb.devel.redhat.com/buildinfo?buildID=371877) comparing with older rhs-hadoop-install-1_33-1.el6rhs.src.rpm (https://brewweb.devel.redhat.com/buildinfo?buildID=371149) In file setup_cluster.sh was previously present "setup_users" function, flags SETUP_LDAP and SETUP_USERS and other stuff related to this. This function is now completely gone and flags used in it too. Arguments --deploy-ldap-srv and --local-auth which setted up mentioned flags are not accepted anymore. Files add_users.sh and add_groups.sh are not there and there is no reference to them. This issue is fixed -> VERIFIED.
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/RHEA-2014-1275.html