Hi Sammi, Just filed https://issues.apache.org/jira/browse/HDDS-14175 . Will submit a pull request soon.
Tsz-wo On Sun, Dec 14, 2025 at 10:49 PM Sammi Chen <[email protected]> wrote: > Thanks Tsz-Wo, HDDS-14164 < > https://issues.apache.org/jira/browse/HDDS-14164> is > merged. > > > - DatanodeDiskBalancerInfoProto should use DatanodeIDProto instead of > DatanodeDetailsProto > > As the disk balancer CLI show hostname instead DN UUID from > returned DatanodeDiskBalancerInfoProto for hostname is more human reader > friendly, and DatanodeIDProto doesn't have hostname, how about introduce a > new DatanodeIDHostProto, so that both UUID and hostname&IP can be returned > for those caller which have only interest of DN address? > > > - In DefaultVolumeChoosingPolicy.chooseVolume(..), getCurrentUsage() is > called multiple times so it can return different values. > Will check this. > > > On Sat, 13 Dec 2025 at 02:54, Tsz Wo Sze <[email protected]> wrote: > > > > - Remove DatanodeDiskBalancerInfoType since it is never used. > > > > It actually is used in DiskBalancerReportSubcommand > > and DiskBalancerStatusSubcommand. However, it is not needed. STATUS is > a > > super set of REPORT -- it includes additional details not included in > > REPORT. I suggest the servers always return all the additional details > in > > order to simplify the code. The commands could choose what to display. > > > > Tsz-Wo > > > > On Fri, Dec 12, 2025 at 10:38 AM Tsz Wo Sze <[email protected]> wrote: > > > > > +1 for the merge although there are some problems listed below. > > > > > > - In DefaultVolumeChoosingPolicy.chooseVolume(..), getCurrentUsage() is > > > called multiple times so it can return different values. The biggest > > > problem is the sorted(..) call. The code calls getCurrentUsage() in > the > > > comparator. The behavior is ill-defined for sorting changing values. > > > > > > - DatanodeDiskBalancerInfoProto should use DatanodeIDProto instead of > > > DatanodeDetailsProto > > > > > > - Remove DatanodeDiskBalancerInfoType since it is never used. > > > > > > - Do not throw RuntimeException in TestContainerChoosingPolicy. How > > about > > > using Assertions.fail(e.getMessage())? > > > > > > Thanks for working on this! > > > Tsz-Wo > > > > > > > > > On Fri, Dec 12, 2025 at 12:49 AM Janus Chow <[email protected]> > > wrote: > > > > > >> +1 for the merge. > > >> > > >> Warm Regards. > > >> > > >> -- > > >> Yiyang Zhou > > >> > > >> > > >> Sadanand Shenoy <[email protected]> 于2025年12月12日周五 13:55写道： > > >> > > >> > +1 for the merge. > > >> > > > >> > - Sadanand > > >> > > > >> > On Wed, Dec 10, 2025 at 3:53 PM Sammi Chen <[email protected]> > > >> wrote: > > >> > > > >> > > Hi Ozone developers, > > >> > > > > >> > > This is for merging the HDDS-5713 Disk Balancer feature branch > into > > >> > master. > > >> > > > > >> > > HDDS-5713 adds the support of the disk volume utilization balancer > > >> > function > > >> > > by selecting and moving containers from high utilized data volume > to > > >> > lower > > >> > > utilized data volume, to achieve an all disk volumes' utilization > > even > > >> > > state. > > >> > > > > >> > > Feature Jira Link: > > >> > > https://issues.apache.org/jira/browse/HDDS-5713 > > >> > > > > >> > > Checklist for feature branch merge: > > >> > > > > >> > > > > >> > > > >> > > > https://cwiki.apache.org/confluence/display/OZONE/Disk+Balancer+For+Datanode+-+HDDS-5713 > > >> > > < > > >> > > > > >> > > > >> > > > https://cwiki.apache.org/confluence/display/OZONE/Disk+Balancer+For+Datanode+-+HDDS-5713 > > >> > > > > > >> > > > > >> > > Please help to take a look and cast the vote if you have time. > > >> > > > > >> > > The vote will be open for at least a week. > > >> > > BTW, previous thread link > > >> > > <https://lists.apache.org/thread/szyx6jxx1f66fzv9bz0t8vq8cdxmt173 > >. > > >> > > > > >> > > Thanks, > > >> > > Sammi Chen > > >> > > > > >> > > > >> > > > > > >
