问题

馆长主失败

T
徽章 ＋３
  • tudamp
  • 资源管理器
  • 0回答
你好,
我有一个备份nutanix集群(CE)的问题。没有虚拟机，它被配置为另一个Nutanix的远端站点。
我们遇到了馆长master的问题，它拒绝从这个错误开始:

代码: 
检查失败:vdisk_config.to_remove() && vdisk_config.has_originating_vdisk_id() && !vdisk_config.has_parent_vdisk_id() &&
_parent_vdisk_id() vdisk_id: 10045 vdisk_size: 0 container_id: 5961 params {} to_remove: true creation_time_usecs: 1512563402344616 mutability_state: kimmutableclosest_named_ancestor: "NFS:4611686018427394362
" vdisk_create_loc: 6 vdisk_create_loc: 3742 vdisk_create_loc: 1110281 nfs_file_name: "c72541c1-9e9b-4b4f-a35a-e8be379e61b7" may_be_parent: true parent_nfs_file_name_hint: "c72541c1-9e9b-4b4f-a35a-e8be379e61
b7" shell_vdisk: true chain_id: "\201\226\377\355{\315L%\245\037{\037\360\231\204\024" vdisk_uuid: " 251\252\027\262\364\377E\000\217Pf%\354\316\323\022" never_hosted: false


我删除了该磁盘(c72541c1-9e9b-4b4f-a35a-e8be379e61b7)的所有快照，它不再存在于存储容器中，但问题仍然存在。

也许问题出在卡桑德拉身上，但我不知道它是如何产生的，也不知道如何解决它。

我可以破坏集群并丢失所有数据，但我想了解之前发生了什么。

有什么建议吗?
谢谢。
\r\nI have a problem with a backup nutanix cluster (CE). There are no VMs, it is configured as a remote site of another Nutanix.
\r\nWe are having problems with the curator master, which refuses to start with this error:
\r\n
\r\n
code:<\/b>
Check failed: vdisk_config.to_remove() && vdisk_config.has_originating_vdisk_id() && !vdisk_config.has_parent_vdisk_id() && !vdisk_config.has_expected
_parent_vdisk_id() vdisk_id: 10045 vdisk_size: 0 container_id: 5961 params { } to_remove: true creation_time_usecs: 1512563402344616 mutability_state: kImmutable closest_named_ancestor: \"NFS:4611686018427394362
\" vdisk_creator_loc: 6 vdisk_creator_loc: 3742 vdisk_creator_loc: 1110281 nfs_file_name: \"c72541c1-9e9b-4b4f-a35a-e8be379e61b7\" may_be_parent: true parent_nfs_file_name_hint: \"c72541c1-9e9b-4b4f-a35a-e8be379e61
b7\" shell_vdisk: true chain_id: \"\\201\\226\\377\\355{\\315L%\\245\\037{\\037\\360\\231\\204\\024\" vdisk_uuid: \"\\251\\252\\027\\262\\364\\377E\\000\\217Pf%\\354\\316\\323\\022\" never_hosted: false
<\/pre><\/div>
\r\n
\r\nI deleted all the snapshots for that disk (c72541c1-9e9b-4b4f-a35a-e8be379e61b7), which is no longer present in the storage container, but the problem persists.
\r\n
\r\nProbably the problem is on cassandra, but I have no idea how it was created and how to solve it. 
\r\n
\r\nI can destroy the cluster and lose all the data, but I'd like to understand what happened before.
\r\n
\r\nAny suggestions?
\r\nThanks.","quoteUsername":"tudamp","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
本主题已关闭供评论

6个回答

aluciani
Userlevel 7
徽章 + 34
你还是有问题 @tudamp？
在Twitter上关注我:https://twitter.com/AngeloLuciani
@tudamp<\/user-mention> ?","quoteUsername":"aluciani","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
T
徽章 ＋３
是的，错误仍然存在
P
排名
Userlevel 3
徽章 ＋４
@tudamp这个问题很可能是由于inconsistent_shell_vdisk_configs，它导致管理员崩溃和/或启动和运行失败(基于附加的代码)。
vdisk_config文件下的一些vdisk可能会将其“shell_vdisk:”标志设置为“True”，而没有设置“originator_vdisk_id”的值。这将导致策展人进入崩溃循环

我们可能需要运行一个修复脚本，该脚本可以纠正不一致性并启动管理器服务。运行该脚本需要支持协助。

上述问题在AOS中得到了解决4.7.4和5.1以上如果AOS版本比这个更老，你可能会遇到同样的问题。

希望这些信息能有所帮助!
@tudamp<\/user-mention> the issue is most likely due to inconsistent_shell_vdisk_configs which is causing the curator to crash and\/or failure to start and run (based on the code that you attached).
\r\nSome vdisks under the vdisk_config file may have their \"shell_vdisk:\" flag set to \"True\" while having no \"originator_vdisk_id\" value set. This causes Curator to enter a crashing loop
\r\n
\r\nWe might need to run a fix script which can correct the inconsistency and bring up the curator service. Running the script would need support assistance.
\r\n
\r\nThe above described issue is resolved in AOS 4.7.4<\/b> and 5.1 above<\/b> and you are likely to hit the same if the AOS version is older to this.
\r\n
\r\n Hope the information helps!","quoteUsername":"Prad_Math","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
T
徽章 ＋３
@Prad_Math谢谢你的信息。
这是一个CE安装，所以我们没有支持。
唯一的解决方案是销毁集群，重新安装/升级到最后的版本?

再次感谢。
@Prad_Math<\/user-mention> thanks for the info.
\r\nThis is a CE installation, so we don't have support.
\r\nThe only solution is to destroy the cluster, re-install\/upgrade to the last version?
\r\n
\r\nThanks again.","quoteUsername":"tudamp","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
aluciani
Userlevel 7
徽章 + 34
对这篇文章的看法 @Primzy @andrew_ct
谢谢
在Twitter上关注我:https://twitter.com/AngeloLuciani
@Primzy<\/user-mention> @andrew_ct<\/user-mention>
\r\nThanks","quoteUsername":"aluciani","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
andrew_ct
Userlevel 5
徽章 + 9
你好,
我在这里只能猜测，但如果您可以毫无损害地破坏集群(取决于您拥有的数据)，我猜直接安装最新版本将是最简单的解决方案，步骤越少越好。
否则我会先尝试修复步骤。
\r\nI can only guess here, but if you can destroy the cluster without harm (depends on the data you have) I guess that directly installing the latest version would be the easiest solution, the fewer steps the better.
\r\nOtherwise I'd try repair steps first.","quoteUsername":"andrew_ct","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
Baidu