Awesome! Nice to hear from you, buddy! MrAsanjar <[email protected]> 於 2020年9月16日 週三 上午3:54寫道：
> Hi Evans, > Let me see what I can do. Give me 24 hr :) > > On Tue, Sep 15, 2020 at 10:51 AM Evans Ye <[email protected]> wrote: > > > Yes. I think the action is correct. However [2] might be a different > thing > > for PPC integration in Hadoop. > > > > Amir, > > Could you confirm? > > > > Kengo Seki <[email protected]> 於 2020年9月14日 週一 下午9:56寫道： > > > >> Thank you for the advice, Evans! > >> Let me confirm about "PPC machine owners". According to Amir's JIRA > >> issues [1][2] and the powered-by list in the OSU site [3], we're using > >> a VM hosted by OSU OSL, right? > >> If it's correct, I'm going to ask them for help via > >> [email protected]. > >> > >> [1]: > >> > https://issues.apache.org/jira/browse/INFRA-11467?focusedCommentId=15300982&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-15300982 > >> [2]: https://issues.apache.org/jira/browse/INFRA-12014 > >> [3]: > https://osuosl.org/services/powerdev/current-projects/#foss-projects > >> > >> Kengo Seki <[email protected]> > >> > >> > >> On Mon, Sep 14, 2020 at 2:06 PM Evans Ye <[email protected]> wrote: > >> > > >> > I'd suggest to reach out to PPC machine owners. Worst case Is we can > >> > temporary drop the PPC support to move the release forward. > >> > > >> > Kengo Seki <[email protected]> 於 2020年9月14日 週一 12:44 寫道： > >> > > >> > > Hi everyone, > >> > > > >> > > Let me share information about the CI environment. > >> > > The worker node for ppc64le is currently offlined, so I just killed > >> all > >> > > jobs > >> > > in the queue waiting for it gets back. Its status is as follows. > >> > > > >> > > - According to the result of `who -b`, that machine seems to be > >> rebooted > >> > > on 2020-09-11 for some reason (probably unexpectedly). > >> > > > >> > > - According to the result of dmesg, the root volume was mounted > >> > > in read-only mode because of a fsck failure. > >> > > > >> > > [ 34.840681] EXT4-fs (vda1): Couldn't remount RDWR because of > >> > > unprocessed orphan inode list. Please umount/remount instead > >> > > [ 60.714110] cgroup: new mount options do not match the existing > >> > > superblock, will be ignored > >> > > [ 316.385805] EXT4-fs (vda1): error count since last fsck: 9459 > >> > > [ 316.385824] EXT4-fs (vda1): initial error at time 1540294049: > >> > > ext4_validate_inode_bitmap:134 > >> > > [ 316.385826] EXT4-fs (vda1): last error at time 1596881526: > >> > > ext4_free_inode:383 > >> > > > >> > > It looks like some fsck work (and replacing the volume, if it fails) > >> > > are required, > >> > > but I'm not sure if I could run something like `e2fsck -p`, because > >> > > I'm also not sure > >> > > where does that machine exist or who's managing it. > >> > > (I slightly thought it was running as a VM with QEMU on some EC2 > >> > > instance, but I couldn't find it) > >> > > > >> > > > Cos, Evans, Olaf > >> > > Would you provide any suggestions? > >> > > > >> > > Kengo Seki <[email protected]> > >> > > > >> > > >
