Hi Lune, May you tell me more about what this method do, please ?
> The postScannerFilterRow is overridden from RegionObserver but it is > actually not implemented in "hbase-agent" so internally it will call Hbase > class only. So it means I need to recompile the whole jar with the modification you advise, right ? > Correct and for the recompilation you can use the basic cmd. Note: Please take backup of the current jar. Regard, Bhavik Patel +91-7208744109 On Wed, Jun 16, 2021 at 6:41 PM lunescar ranger <lunesc...@gmail.com> wrote: > Hello Patel. > > Thanks for this first answer. > > May you tell me more about what this method do, please ? > > I can find the class you mentioned in the following jar : > ranger-hbase-plugin-0.7.0.2.6.2.0-205.jar > > So it means I need to recompile the whole jar with the modification you > advise, right ? > As I use the complete stack HDP 2.6.2 of Cloudera, do you know if there is > some documentation to do this recompilation (which dependencies to use for > example) ? > > Best regards. > > Lune > > > Le mar. 15 juin 2021 à 16:49, Bhavik Patel <bhavikpatel...@gmail.com> a > écrit : > >> Hi Lune, >> >> Can you try by removing the postScannerFilterRow from the >> RangerAuthorizationCoprocessor class( >> https://github.com/apache/ranger/blob/master/ranger-hbase-plugin-shim/src/main/java/org/apache/ranger/authorization/hbase/RangerAuthorizationCoprocessor.java#L1772 >> ). >> >> Please let me know if it works for you. >> >> >> Thanks, >> Bhavik Patel >> +91-7208744109 >> >> >> On Tue, Jun 15, 2021, 7:19 PM lunescar ranger <lunesc...@gmail.com> >> wrote: >> >>> Hello ! >>> >>> I send you this mail because I have a problem when I try to enable the >>> Ranger plugin for HBase. >>> I have an HDP 2.6.2 (HBase version is 1.1.2) with 500 HBase >>> regionservers and 2 HBase Masters. >>> I have around 100 000 regions in this HBase cluster. >>> >>> Without Ranger, when I start the cluster, it takes around 20-30 minuts >>> to be started, and then it answers normally. >>> >>> With Ranger enabled for HBase, when I start the cluster, it takes more >>> than an hour to start, and then when I query HBase to scan or to put, I >>> often encounter timeout errors from the regionservers. >>> >>> Is it a known issue that Ranger provoke a performance drop in HBase ? >>> Is there a way to solve this performance problem ? >>> >>> Thank you in advance for your answers. >>> >>> Best regards. >>> >>> Lune >>> >>
