Alexey Serbin has posted comments on this change. Change subject: minicluster: facilitate creating multidir clusters ......................................................................
Patch Set 9: (9 comments) http://gerrit.cloudera.org:8080/#/c/6845/9/src/kudu/integration-tests/external_mini_cluster.cc File src/kudu/integration-tests/external_mini_cluster.cc: PS9, Line 255: dir_nums nit: not sure that's the best name for the parameter what about 'dir_suffix' or alike? http://gerrit.cloudera.org:8080/#/c/6845/9/src/kudu/integration-tests/external_mini_cluster.h File src/kudu/integration-tests/external_mini_cluster.h: PS9, Line 26: #include <sys/types.h> This is from the system C headers, this should come the very first in this file. http://gerrit.cloudera.org:8080/#/c/6845/9/src/kudu/integration-tests/multidisk-cluster-itest.cc File src/kudu/integration-tests/multidisk-cluster-itest.cc: Line 21: nit: consider adding #include <gtest/gtest.h> PS9, Line 26: #include "kudu/integration-tests/test_workload.h" I don't think it's necessary. PS9, Line 44: TabletServerIntegrationTestBase Why not just ExternalMiniClusterITestBase ? Is there anything specific you want to bring from tablet-server specific test? PS9, Line 46: MultiDiskClusterITest() {} I think you can drop this empty constructor -- it will be automatically generated for you by the compiler. PS9, Line 93: CHECK_GT why not ASSERT_GT()? What about just ASSERT_FALSE(tablet_replicas_.empty()); PS9, Line 129: ASSERT_GT(1, num_dirs_added_to); I'm confused by the combination of the comment and this assert: if I'm not mistaken, the latter is equivalent to ASSERT_TRUE(1 > num_dirs_added_to) and the former contradicts with that. Consider either clarifying on the comment or updating the assert condition. Or I'm missing something :) http://gerrit.cloudera.org:8080/#/c/6845/9/src/kudu/util/path_util.cc File src/kudu/util/path_util.cc: PS9, Line 58: string ListPaths(const vector<string>& paths) { : if (paths.empty()) return ""; : std::stringstream ss; : for (int path_num = 0; path_num < paths.size(); path_num++) { : if (path_num != 0) { : ss << ","; : } : ss << paths[path_num]; : } : return ss.str(); : } It seems using JoinStrings(paths, ",") could be a better fit here. If so, does it make sense to bring this new function at all? If switching to JoinStrings, don't forget to remove the <sstream> include. -- To view, visit http://gerrit.cloudera.org:8080/6845 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: Id2f5def6980ad394c8558ad97ba830f1b0257332 Gerrit-PatchSet: 9 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: Adar Dembo <a...@cloudera.com> Gerrit-Reviewer: Alexey Serbin <aser...@cloudera.com> Gerrit-Reviewer: Andrew Wong <aw...@cloudera.com> Gerrit-Reviewer: David Ribeiro Alves <davidral...@gmail.com> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <mpe...@apache.org> Gerrit-Reviewer: Tidy Bot Gerrit-HasComments: Yes
