You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ozone.apache.org by Sammi Chen <sa...@apache.org> on 2023/11/14 09:36:13 UTC
Ozone Community Meeting(APAC, Nov 10th)
Attender: Xi, Hao, Minyu, Bin, Yuanben, Guangbao, Jianghua, Conway, Sammi
Agenda:
-
Qihoo
1. Delivered a great short session about the learning and customization
of Ozone Manager.
OM改造分享-Qihoo.pdf
<https://github.com/apache/ozone/files/13347053/OM.-Qihoo.pdf>
-
DiDi
1. Using a dedicated cluster to reproduce and investigate how to improve
the OM IOPS (write and read). Previous cluster can achieve
4K+4K(createKey,commitKey) IOPS.
2. Is suggested to evaluate HDDS-9279
<https://issues.apache.org/jira/browse/HDDS-9279> OM HA: support read
from followers.
Re: Ozone Community Meeting(APAC, Nov 10th)
Posted by Tsz Wo Sze <sz...@gmail.com>.
> Is suggested to evaluate HDDS-9279 ...
Thanks a lot for checking HDDS-9279! The pull request is basically working
and ready for evaluation. The main missing part is that if a client fails
to talk to a server, it should failover to another server. Please keep me
posted and feel free to let me know if there is anything I could help with.
Tsz-Wo
On Tue, Nov 14, 2023 at 1:36 AM Sammi Chen <sa...@apache.org> wrote:
> Attender: Xi, Hao, Minyu, Bin, Yuanben, Guangbao, Jianghua, Conway, Sammi
> Agenda:
>
> -
>
> Qihoo
> 1. Delivered a great short session about the learning and customization
> of Ozone Manager.
> OM改造分享-Qihoo.pdf
> <https://github.com/apache/ozone/files/13347053/OM.-Qihoo.pdf>
> -
>
> DiDi
> 1. Using a dedicated cluster to reproduce and investigate how to improve
> the OM IOPS (write and read). Previous cluster can achieve
> 4K+4K(createKey,commitKey) IOPS.
> 2. Is suggested to evaluate HDDS-9279
> <https://issues.apache.org/jira/browse/HDDS-9279> OM HA: support
> read
> from followers.
>
Re: Ozone Community Meeting(APAC, Nov 10th)
Posted by Tsz Wo Sze <sz...@gmail.com>.
> Is suggested to evaluate HDDS-9279 ...
Thanks a lot for checking HDDS-9279! The pull request is basically working
and ready for evaluation. The main missing part is that if a client fails
to talk to a server, it should failover to another server. Please keep me
posted and feel free to let me know if there is anything I could help with.
Tsz-Wo
On Tue, Nov 14, 2023 at 1:36 AM Sammi Chen <sa...@apache.org> wrote:
> Attender: Xi, Hao, Minyu, Bin, Yuanben, Guangbao, Jianghua, Conway, Sammi
> Agenda:
>
> -
>
> Qihoo
> 1. Delivered a great short session about the learning and customization
> of Ozone Manager.
> OM改造分享-Qihoo.pdf
> <https://github.com/apache/ozone/files/13347053/OM.-Qihoo.pdf>
> -
>
> DiDi
> 1. Using a dedicated cluster to reproduce and investigate how to improve
> the OM IOPS (write and read). Previous cluster can achieve
> 4K+4K(createKey,commitKey) IOPS.
> 2. Is suggested to evaluate HDDS-9279
> <https://issues.apache.org/jira/browse/HDDS-9279> OM HA: support
> read
> from followers.
>