Mike Percy has posted comments on this change. ( http://gerrit.cloudera.org:8080/8778 )
Change subject: docs: updates to data dir behavior ...................................................................... Patch Set 3: (11 comments) looks good. added a few suggestions / thoughts, lmk what you think http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc File docs/administration.adoc: http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@731 PS3, Line 731: the server must be emptied the wal and data directories on the server must be deleted http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@732 PS3, Line 732: . , destroying the local copy of the data for each tablet replica hosted on the local server. http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@733 PS3, Line 733: . if the replication factor is at least 3 and all other tablet servers are online and healthy http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@738 PS3, Line 738: server this will work on masters too, right? might want to add a note that these are instructions for tablet servers but the process is similar for master servers as well http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@747 PS3, Line 747: rm -rf /data/0/kudu-tserver-wal/* /data/1/kudu-tserver/* /data/2/kudu-tserver/* add comment: # Note: this will delete all of the data from the local tablet server http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@755 PS3, Line 755: emptied deleted http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@756 PS3, Line 756: created with the appropriate permissions You didn't mention this. shouldn't this happen automatically? Perhaps we should mention that it's automatic. How can they tell whether it worked? http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@762 PS3, Line 762: All directories on Kudu masters and select directories (the WAL directory and : the metadata directory) of tablet servers are not resilient to disk failures. How about: As of Kudu 1.6.0, Kudu master servers are not resilient to any types of disk failures. Kudu tablet servers are only resilient to disk failures if they occur on a disk storing data blocks, so the failure of a disk where the write-ahead logs or tablet metadata are stored will still result in data loss of the entire tablet server. http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@764 PS3, Line 764: these non-resilient http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@769 PS3, Line 769: failures of the other directories maybe: a disk failure on one of the failure-resilient directories (?) http://gerrit.cloudera.org:8080/#/c/8778/3/docs/administration.adoc@769 PS3, Line 769: automatically : re-replicate tablets with data on the affected disk, and the affected server : will remain alive with the logs indicating the failure, for example: : automatically stop using the affected disk, shut down tablets with blocks on the affected disk, and re-replicate the affected tablets to other tablet servers. The affected server will remain alive and will print messages to the log indicating the disk failure, for example: -- To view, visit http://gerrit.cloudera.org:8080/8778 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I7cfef4aeaba92228d2e0a77c7596847a6a3137e3 Gerrit-Change-Number: 8778 Gerrit-PatchSet: 3 Gerrit-Owner: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <mpe...@apache.org> Gerrit-Reviewer: Will Berkeley <wdberke...@gmail.com> Gerrit-Comment-Date: Thu, 07 Dec 2017 20:33:42 +0000 Gerrit-HasComments: Yes