[lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[GEORGE ZHAO]

[-- Attachment #1.1: Type: text/plain, Size: 597 bytes --]

Hi Folks ,

I'm a storage engineer and a Lustre beginner who just figured out how to
deploy it. Did read a lot about it though.
My goal is using Lustre to support HPC/ML workload. And I believe doing
some tasks would help me proceed better.

As wiki suggested, I found an "easy" ticket LU-1643: “Request to remove
display of reserved space from statfs (df) total space.”
Is it still unclaimed?
If it's already in progress, or there's a higher priority issue that could
use help, please point me in the right direction.

Thanks in advance for any advice.

Best regards,
George

[-- Attachment #1.2: Type: text/html, Size: 746 bytes --]

[-- Attachment #2: Type: text/plain, Size: 165 bytes --]

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Re: [lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[Peter Jones]

[-- Attachment #1.1: Type: text/plain, Size: 1349 bytes --]

George

Thanks for your interest in Lustre! This list - https://jira.whamcloud.com/issues/?filter=10990 – should provide a wide range of options.  The convention is that anyone working on a ticket assigns it to themselves to prevent duplicate effort

Peter

From: lustre-devel <lustre-devel-bounces@lists.lustre.org> on behalf of GEORGE ZHAO <georgezhaojobs@gmail.com>
Date: Tuesday, April 29, 2025 at 4:42 PM
To: "lustre-devel@lists.lustre.org" <lustre-devel@lists.lustre.org>
Subject: [lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

You don't often get email from georgezhaojobs@gmail.com. Learn why this is important<https://aka.ms/LearnAboutSenderIdentification>
Hi Folks ,

I'm a storage engineer and a Lustre beginner who just figured out how to deploy it. Did read a lot about it though.
My goal is using Lustre to support HPC/ML workload. And I believe doing some tasks would help me proceed better.

As wiki suggested, I found an "easy" ticket LU-1643: “Request to remove display of reserved space from statfs (df) total space.”
Is it still unclaimed?
If it's already in progress, or there's a higher priority issue that could use help, please point me in the right direction.

Thanks in advance for any advice.

Best regards,
George

[-- Attachment #1.2: Type: text/html, Size: 11878 bytes --]

[-- Attachment #2: Type: text/plain, Size: 165 bytes --]

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Re: [lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[Andreas Dilger]

[-- Attachment #1.1: Type: text/plain, Size: 2530 bytes --]

Hi George,
thanks for your interest.  There are definitely a number of open tickets in the "easy" (and "medium") categories that you could work on, though there have been a number of interns and new hires working on those tickets recently and fewer are now available.

LU-1643 happens to be the oldest "easy" ticket, but don't let that guide your selection.  I would suggest you take a look through the whole list and pick something that is of interest to you and/or matches your experience level.

- LU-16738<https://jira.whamcloud.com/browse/LU-16738> - Improve mount.lustre with many MGS NIDs<https://jira.whamcloud.com/browse/LU-16738>: this is partly a userspace (mount.lustre) change and partly in the kernel (though not incredibly complex).  This is actually in the "useful" category that provides improved functionality for something that is seen regularly.


If you are feeling more adventurous, you might look at one of the "medium" tickets, which may not be too difficult:
- LU-7880<https://jira.whamcloud.com/browse/LU-7880> - add OST/MDT performance statistics to obd_statfs<https://jira.whamcloud.com/browse/LU-7880>:  this is a very useful feature that would allow clients/user applications to probe the IO load on OSTs, and self-optimize checkpoint IO to times when the OSTs are less busy
- LU-13123<https://jira.whamcloud.com/browse/LU-13123> - Add list of client NIDs to job_stats output<https://jira.whamcloud.com/browse/LU-13123>: this would associate all of the clients that are contributing to the job_stats for a specific JobID, and would be useful for debugging and monitoring/

On Apr 29, 2025, at 17:41, GEORGE ZHAO <georgezhaojobs@gmail.com> wrote:

I'm a storage engineer and a Lustre beginner who just figured out how to deploy it. Did read a lot about it though.
My goal is using Lustre to support HPC/ML workload. And I believe doing some tasks would help me proceed better.

As wiki suggested, I found an "easy" ticket LU-1643: “Request to remove display of reserved space from statfs (df) total space.”
Is it still unclaimed?
If it's already in progress, or there's a higher priority issue that could use help, please point me in the right direction.

Thanks in advance for any advice.

Best regards,
George
_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Cheers, Andreas
—
Andreas Dilger
Lustre Principal Architect
Whamcloud/DDN





[-- Attachment #1.2: Type: text/html, Size: 4627 bytes --]

[-- Attachment #2: Type: text/plain, Size: 165 bytes --]

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Re: [lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[GEORGE ZHAO]

[-- Attachment #1.1: Type: text/plain, Size: 3045 bytes --]

Thank you, Andreas!
I'm currently looking into LU-7880. Will follow up if I have an
implementation plan or run into any questions (quite likely).

By the way, should I reach out to the lustre-devel for questions like this,
or is there a more specific channel?

Thanks!
George

On Wed, Apr 30, 2025 at 4:44 AM Andreas Dilger <adilger@ddn.com> wrote:

> Hi George,
> thanks for your interest.  There are definitely a number of open tickets
> in the "easy" (and "medium") categories that you could work on, though
> there have been a number of interns and new hires working on those tickets
> recently and fewer are now available.
>
> LU-1643 happens to be the oldest "easy" ticket, but don't let that guide
> your selection.  I would suggest you take a look through the whole list and
> pick something that is of interest to you and/or matches your experience
> level.
>
> - LU-16738 <https://jira.whamcloud.com/browse/LU-16738> - Improve
> mount.lustre with many MGS NIDs
> <https://jira.whamcloud.com/browse/LU-16738>: this is partly a userspace
> (mount.lustre) change and partly in the kernel (though not incredibly
> complex).  This is actually in the "useful" category that provides improved
> functionality for something that is seen regularly.
>
>
> If you are feeling more adventurous, you might look at one of the "medium"
> tickets, which may not be too difficult:
> - LU-7880 <https://jira.whamcloud.com/browse/LU-7880> - add OST/MDT
> performance statistics to obd_statfs
> <https://jira.whamcloud.com/browse/LU-7880>:  this is a very useful
> feature that would allow clients/user applications to probe the IO load on
> OSTs, and self-optimize checkpoint IO to times when the OSTs are less busy
> - LU-13123 <https://jira.whamcloud.com/browse/LU-13123> - Add list of
> client NIDs to job_stats output
> <https://jira.whamcloud.com/browse/LU-13123>: this would associate all of
> the clients that are contributing to the job_stats for a specific JobID,
> and would be useful for debugging and monitoring/
>
> On Apr 29, 2025, at 17:41, GEORGE ZHAO <georgezhaojobs@gmail.com> wrote:
>
>
> I'm a storage engineer and a Lustre beginner who just figured out how to
> deploy it. Did read a lot about it though.
> My goal is using Lustre to support HPC/ML workload. And I believe doing
> some tasks would help me proceed better.
>
> As wiki suggested, I found an "easy" ticket LU-1643: “Request to remove
> display of reserved space from statfs (df) total space.”
> Is it still unclaimed?
> If it's already in progress, or there's a higher priority issue that could
> use help, please point me in the right direction.
>
> Thanks in advance for any advice.
>
> Best regards,
> George
> _______________________________________________
> lustre-devel mailing list
> lustre-devel@lists.lustre.org
> http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org
>
>
> Cheers, Andreas
> —
> Andreas Dilger
> Lustre Principal Architect
> Whamcloud/DDN
>
>
>
>
>

[-- Attachment #1.2: Type: text/html, Size: 4779 bytes --]

[-- Attachment #2: Type: text/plain, Size: 165 bytes --]

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Re: [lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[Andreas Dilger]

[-- Attachment #1.1: Type: text/plain, Size: 3253 bytes --]

There is a Slack channel, which is linked from wiki.lustre.org "lists" page that you can use for simple discussions. Email is probably better for complex questions.

Cheers, Andreas

On Apr 30, 2025, at 20:18, GEORGE ZHAO <georgezhaojobs@gmail.com> wrote:


Thank you, Andreas!
I'm currently looking into LU-7880. Will follow up if I have an implementation plan or run into any questions (quite likely).

By the way, should I reach out to the lustre-devel for questions like this, or is there a more specific channel?

Thanks!
George

On Wed, Apr 30, 2025 at 4:44 AM Andreas Dilger <adilger@ddn.com<mailto:adilger@ddn.com>> wrote:
Hi George,
thanks for your interest.  There are definitely a number of open tickets in the "easy" (and "medium") categories that you could work on, though there have been a number of interns and new hires working on those tickets recently and fewer are now available.

LU-1643 happens to be the oldest "easy" ticket, but don't let that guide your selection.  I would suggest you take a look through the whole list and pick something that is of interest to you and/or matches your experience level.

- LU-16738<https://jira.whamcloud.com/browse/LU-16738> - Improve mount.lustre with many MGS NIDs<https://jira.whamcloud.com/browse/LU-16738>: this is partly a userspace (mount.lustre) change and partly in the kernel (though not incredibly complex).  This is actually in the "useful" category that provides improved functionality for something that is seen regularly.


If you are feeling more adventurous, you might look at one of the "medium" tickets, which may not be too difficult:
- LU-7880<https://jira.whamcloud.com/browse/LU-7880> - add OST/MDT performance statistics to obd_statfs<https://jira.whamcloud.com/browse/LU-7880>:  this is a very useful feature that would allow clients/user applications to probe the IO load on OSTs, and self-optimize checkpoint IO to times when the OSTs are less busy
- LU-13123<https://jira.whamcloud.com/browse/LU-13123> - Add list of client NIDs to job_stats output<https://jira.whamcloud.com/browse/LU-13123>: this would associate all of the clients that are contributing to the job_stats for a specific JobID, and would be useful for debugging and monitoring/

On Apr 29, 2025, at 17:41, GEORGE ZHAO <georgezhaojobs@gmail.com<mailto:georgezhaojobs@gmail.com>> wrote:

I'm a storage engineer and a Lustre beginner who just figured out how to deploy it. Did read a lot about it though.
My goal is using Lustre to support HPC/ML workload. And I believe doing some tasks would help me proceed better.

As wiki suggested, I found an "easy" ticket LU-1643: “Request to remove display of reserved space from statfs (df) total space.”
Is it still unclaimed?
If it's already in progress, or there's a higher priority issue that could use help, please point me in the right direction.

Thanks in advance for any advice.

Best regards,
George
_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org<mailto:lustre-devel@lists.lustre.org>
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Cheers, Andreas
—
Andreas Dilger
Lustre Principal Architect
Whamcloud/DDN





[-- Attachment #1.2: Type: text/html, Size: 5479 bytes --]

[-- Attachment #2: Type: text/plain, Size: 165 bytes --]

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org

Re: [lustre-devel] New contributor inquiry – status of LU-1643 (“reserved space” in statfs/df) and guidance requested

[GEORGE ZHAO]

[-- Attachment #1.1: Type: text/plain, Size: 3836 bytes --]

Hi Andreas,

Could you or Simmons send me an invitation to this channel?
https://lustreupstreamclient.slack.com/messages/CAV4QF0ER/
An invitation from the workspace admin is required (James Simmons)

I have a question about how to retrieve data from /proc.
And I'd like to clarify the behavior of "lfs df --stats".

Thanks,
George

On Thu, May 1, 2025 at 1:52 AM Andreas Dilger <adilger@ddn.com> wrote:

> There is a Slack channel, which is linked from wiki.lustre.org "lists"
> page that you can use for simple discussions. Email is probably better for
> complex questions.
>
> Cheers, Andreas
>
> On Apr 30, 2025, at 20:18, GEORGE ZHAO <georgezhaojobs@gmail.com> wrote:
>
> 
> Thank you, Andreas!
> I'm currently looking into LU-7880. Will follow up if I have an
> implementation plan or run into any questions (quite likely).
>
> By the way, should I reach out to the lustre-devel for questions like
> this, or is there a more specific channel?
>
> Thanks!
> George
>
> On Wed, Apr 30, 2025 at 4:44 AM Andreas Dilger <adilger@ddn.com> wrote:
>
>> Hi George,
>> thanks for your interest.  There are definitely a number of open tickets
>> in the "easy" (and "medium") categories that you could work on, though
>> there have been a number of interns and new hires working on those tickets
>> recently and fewer are now available.
>>
>> LU-1643 happens to be the oldest "easy" ticket, but don't let that guide
>> your selection.  I would suggest you take a look through the whole list and
>> pick something that is of interest to you and/or matches your experience
>> level.
>>
>> - LU-16738 <https://jira.whamcloud.com/browse/LU-16738> - Improve
>> mount.lustre with many MGS NIDs
>> <https://jira.whamcloud.com/browse/LU-16738>: this is partly a userspace
>> (mount.lustre) change and partly in the kernel (though not incredibly
>> complex).  This is actually in the "useful" category that provides improved
>> functionality for something that is seen regularly.
>>
>>
>> If you are feeling more adventurous, you might look at one of the
>> "medium" tickets, which may not be too difficult:
>> - LU-7880 <https://jira.whamcloud.com/browse/LU-7880> - add OST/MDT
>> performance statistics to obd_statfs
>> <https://jira.whamcloud.com/browse/LU-7880>:  this is a very useful
>> feature that would allow clients/user applications to probe the IO load on
>> OSTs, and self-optimize checkpoint IO to times when the OSTs are less busy
>> - LU-13123 <https://jira.whamcloud.com/browse/LU-13123> - Add list of
>> client NIDs to job_stats output
>> <https://jira.whamcloud.com/browse/LU-13123>: this would associate all
>> of the clients that are contributing to the job_stats for a specific JobID,
>> and would be useful for debugging and monitoring/
>>
>> On Apr 29, 2025, at 17:41, GEORGE ZHAO <georgezhaojobs@gmail.com> wrote:
>>
>>
>> I'm a storage engineer and a Lustre beginner who just figured out how to
>> deploy it. Did read a lot about it though.
>> My goal is using Lustre to support HPC/ML workload. And I believe doing
>> some tasks would help me proceed better.
>>
>> As wiki suggested, I found an "easy" ticket LU-1643: “Request to remove
>> display of reserved space from statfs (df) total space.”
>> Is it still unclaimed?
>> If it's already in progress, or there's a higher priority issue that
>> could use help, please point me in the right direction.
>>
>> Thanks in advance for any advice.
>>
>> Best regards,
>> George
>> _______________________________________________
>> lustre-devel mailing list
>> lustre-devel@lists.lustre.org
>> http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org
>>
>>
>> Cheers, Andreas
>> —
>> Andreas Dilger
>> Lustre Principal Architect
>> Whamcloud/DDN
>>
>>
>>
>>
>>

[-- Attachment #1.2: Type: text/html, Size: 6420 bytes --]

[-- Attachment #2: Type: text/plain, Size: 165 bytes --]

_______________________________________________
lustre-devel mailing list
lustre-devel@lists.lustre.org
http://lists.lustre.org/listinfo.cgi/lustre-devel-lustre.org