I believe there is some misstep on my part. I think I neglected to stop or kill drillbit before installing or upgrading new rpms.
Sungwook On Tue, Dec 30, 2014 at 8:33 AM, Andries Engelbrecht < [email protected]> wrote: > The upgrade from 0.6r2 to 0.7 seemed to be a bit more smooth then > previously. > > I have only performed it on CentOS/RH using yum upgrade. > > It did update the warden.drill-bits.conf file to point to the new drill > version and MCS seems to recognize the new sw just fine, so I never did the > configure.sh -R step. Is the configure step actually required? > > > I did observe that the S3 SP needed to be reconfigured and also some > changes in the drill/conf dir and files were not carried over. At least the > views seemed to have survived the upgrade this time. > > > —Andries > > > > > > On Dec 30, 2014, at 7:36 AM, Sungwook Yoon <[email protected]> wrote: > > > Hi, > > > > According to MapR's document on drill upgrade, > > http://doc.mapr.com/display/MapR/Upgrading+Apache+Drill > > > > After stop, update rpm then, configure.sh -R on the cluster nodes. > > I see these messages, > > > > grep: /opt/mapr/drill/drill-0.7.0/conf/drill-env.sh: No such file or > > directory > > /opt/mapr/server/configure.sh: line 1088: > > /opt/mapr/drill/drill-0.7.0/conf/drill-env.sh: No such file or directory > > /opt/mapr/server/configure.sh: line 1089: > > /opt/mapr/drill/drill-0.7.0/conf/drill-env.sh: No such file or directory > > > > Is this normal? > > > > Thanks, > > > > Sungwook > >
