<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
</head>
<body>
<div dir="ltr">
<div style=""></div>
<div style="">
<div dir="ltr">Hi Christoph,</div>
<div dir="ltr"><br>
</div>
<div dir="ltr">We can start IO once first AFM snapshot (psnap0) is created at the primary after starting changeSecondary command. We might get psnap0 creation failure with heavy workload.</div>
<div dir="ltr"><br>
</div>
<div dir="ltr">This might not be needed if we have fileset level quiesce feature.</div>
<div dir="ltr"><br>
</div>
<div dir="ltr"><br>
</div>
<div dir="ltr"><br>
</div>
<div dir="ltr"><br>
</div>
<div dir="ltr">Thanks & Regards,</div>
<div dir="ltr">Ashish </div>
</div>
<div id="mail-editor-reference-message-container" class="ms-outlook-mobile-reference-message">
<hr style="display:inline-block;width:98%" tabindex="-1">
<div id="divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif"><b>From:</b> gpfsug-discuss <gpfsug-discuss-bounces@gpfsug.org> on behalf of gpfsug-discuss-request@gpfsug.org <gpfsug-discuss-request@gpfsug.org><br>
<b>Sent:</b> Friday, March 31, 2023 4:30 PM<br>
<b>To:</b> gpfsug-discuss@gpfsug.org <gpfsug-discuss@gpfsug.org><br>
<b>Subject:</b> [EXTERNAL] gpfsug-discuss Digest, Vol 132, Issue 25
<div> </div>
</font></div>
<meta name="Generator" content="Microsoft Exchange Server">
<!-- converted from text --><font size="2"><span style="font-size:11pt;">
<div class="PlainText">Send gpfsug-discuss mailing list submissions to<br>
gpfsug-discuss@gpfsug.org<br>
<br>
To subscribe or unsubscribe via the World Wide Web, visit<br>
<a href="http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org">http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org</a>
<br>
or, via email, send a message with subject or body 'help' to<br>
gpfsug-discuss-request@gpfsug.org<br>
<br>
You can reach the person managing the list at<br>
gpfsug-discuss-owner@gpfsug.org<br>
<br>
When replying, please edit your Subject line so it is more specific<br>
than "Re: Contents of gpfsug-discuss digest..."<br>
<br>
<br>
Today's Topics:<br>
<br>
1. AFM convertToPrimary (Christoph Martin)<br>
2. Call for Submissions IO500 ISC23 (IO500 Committee)<br>
<br>
<br>
----------------------------------------------------------------------<br>
<br>
Message: 1<br>
Date: Thu, 30 Mar 2023 18:44:18 +0200<br>
From: Christoph Martin <martin@uni-mainz.de><br>
To: gpfsug main discussion list <gpfsug-discuss@gpfsug.org><br>
Subject: [gpfsug-discuss] AFM convertToPrimary<br>
Message-ID: <436de125-ffed-3d10-bdea-1eb005923964@uni-mainz.de><br>
Content-Type: text/plain; charset="utf-8"; Format="flowed"<br>
<br>
Hi all,<br>
<br>
we want to convert two GPFS independent fileset to primaries for AFM-DR.<br>
Most of the data from fileset1 we have already mirrored via rsync to the <br>
secondary fileset. fileset2 on the secondary server is still empty.<br>
<br>
I now want to issue for both filesets something like:<br>
<br>
mmafmctl storage1 convertToPrimary -j afmdrtest2 --afmtarget <br>
gpfs:///gpfs/storage2/afmdrtest2/ --inband --check-metadata<br>
<br>
The help pages say, that you have to stop all IO on the primary fileset <br>
for the conversion. I think this might mean read and write IO.<br>
<br>
The question is: When can we restart IO? After the mmafmctl command <br>
finishes or after the synchronisation of the filesets finishes.<br>
<br>
The second can take days, especially for the second fileset, which is <br>
empty on the secondary site.<br>
<br>
What happens, if read (or write) IO happens during the execution time of <br>
mmafmctl or the synchronisation?<br>
<br>
Any ideas?<br>
<br>
Regards<br>
Christoph<br>
-------------- next part --------------<br>
A non-text attachment was scrubbed...<br>
Name: OpenPGP_signature<br>
Type: application/pgp-signature<br>
Size: 833 bytes<br>
Desc: OpenPGP digital signature<br>
URL: <<a href=""></a>http://gpfsug.org/pipermail/gpfsug-discuss_gpfsug.org/attachments/20230330/11417e0a/attachment-0001.sig
><br>
<br>
------------------------------<br>
<br>
Message: 2<br>
Date: Thu, 30 Mar 2023 17:48:11 -0600<br>
From: IO500 Committee <committee@io500.org><br>
To: BeeGFS <fhgfs-user@googlegroups.com>, Ceph <ceph-users@ceph.io>,<br>
GPFS <gpfsug-discuss@spectrumscale.org>, HPC Announce<br>
<hpc-announce@mcs.anl.gov>, IO500 <io-500@vi4io.org>, Lustre<br>
<lustre-discuss@lists.lustre.org>, OrangeFS<br>
<users@lists.orangefs.org>, Storage Research List<br>
<storage-research-list@ece.cmu.edu><br>
Subject: [gpfsug-discuss] Call for Submissions IO500 ISC23<br>
Message-ID: <1ceb433e1ae046c5dcd515a69a807e26@io500.org><br>
Content-Type: text/plain; charset=US-ASCII; format=flowed<br>
<br>
Stabilization Period: Monday, April 3rd - Friday, April 14th, 2023<br>
Submission Deadline: Tuesday, May 16st, 2023 AoE<br>
<br>
The IO500 is now accepting and encouraging submissions for the upcoming <br>
12th semi-annual IO500 list, in conjunction with ISC23. Once again, we <br>
are also accepting submissions to the 10 Client Node Challenge to <br>
encourage the submission of small scale results. The new ranked lists <br>
will be announced at the ISC23 BoF [1]. We hope to see many new <br>
results.<br>
<br>
What's New<br>
1. Creation of Production and Research Lists - Starting with ISC'22, we <br>
proposed a separation of the list into separate Production and Research <br>
lists. This better reflects the important distinction between storage <br>
systems that run in production environments and those that may use more <br>
experimental hardware and software configurations. At ISC23, we will <br>
formally create these two lists and users will be able to submit to <br>
either of the two lists (and their 10 client-node counterparts). Please <br>
see the requirements for each list on the IO500 rules page [3].<br>
2. New Submission Tool - There is now a new IO500 submission tool that <br>
improves the overall submission experience. Users can create accounts <br>
and then update and manage all of their submissions through that <br>
account. As part of this new tool, we have improved the submission <br>
fields that describe the hardware and software of the system under test. <br>
For reproducibility and analysis reasons, we now made the easily <br>
obtainable fields mandatory - data from storage servers are for users <br>
often difficult to obtain, therefore, most remain optional. As a new <br>
system, there may be quirks, please reach out on Slack or the mailing <br>
list if you see any issues. Further details will be released on the <br>
submission page [2].<br>
3. Reproducibility - Every submission will now receive a reproducibility <br>
score based upon the provided system details and the reproducibility <br>
questionnaire. This score will inform the community on the amount of <br>
details provided in the submission and the obtainability of the storage <br>
system. Further, this score will be used to evaluate if a submission is <br>
eligible for the Production list.<br>
4. New Phases - We are continuing to evaluate the inclusion of optional <br>
test phases for additional key workloads - split easy/hard find phases, <br>
4KB and 1MB random read/write phases, and concurrent metadata <br>
operations. This is called an extended run. At the moment, we collect <br>
the information to verify that additional phases do not significantly <br>
impact the results of a standard run and an extended run to facilitate <br>
comparisons between the existing and new benchmark phases. In a future <br>
release, we may include some or all of these results as part of the <br>
standard benchmark. The extended results are not currently included in <br>
the scoring of any ranked list.<br>
<br>
Background<br>
The benchmark suite is designed to be easy to run and the community has <br>
multiple active support channels to help with any questions. Please note <br>
that submissions of all sizes are welcome; the site has customizable <br>
sorting, so it is possible to submit on a small system and still get a <br>
very good per-client score, for example. Additionally, the list is about <br>
much more than just the raw rank; all submissions help the community by <br>
collecting and publishing a wider corpus of data. More details below.<br>
<br>
Following the success of the Top500 in collecting and analyzing <br>
historical trends in supercomputer technology and evolution, the IO500 <br>
was created in 2017, published its first list at SC17, and has grown <br>
continually since then. The need for such an initiative has long been <br>
known within High-Performance Computing; however, defining appropriate <br>
benchmarks has long been challenging. Despite this challenge, the <br>
community, after long and spirited discussion, finally reached consensus <br>
on a suite of benchmarks and a metric for resolving the scores into a <br>
single ranking.<br>
<br>
The multi-fold goals of the benchmark suite are as follows:<br>
1. Maximizing simplicity in running the benchmark suite<br>
2. Encouraging optimization and documentation of tuning parameters for <br>
performance<br>
3. Allowing submitters to highlight their "hero run" performance numbers<br>
4. Forcing submitters to simultaneously report performance for <br>
challenging IO patterns.<br>
<br>
Specifically, the benchmark suite includes a hero-run of both IOR and <br>
mdtest configured however possible to maximize performance and establish <br>
an upper-bound for performance. It also includes an IOR and mdtest run <br>
with highly prescribed parameters in an attempt to determine a lower <br>
performance bound. Finally, it includes a namespace search as this has <br>
been determined to be a highly sought-after feature in HPC storage <br>
systems that has historically not been well-measured. Submitters are <br>
encouraged to share their tuning insights for publication.<br>
<br>
The goals of the community are also multi-fold:<br>
1. Gather historical data for the sake of analysis and to aid <br>
predictions of storage futures<br>
2. Collect tuning information to share valuable performance <br>
optimizations across the community<br>
3. Encourage vendors and designers to optimize for workloads beyond <br>
"hero runs"<br>
4. Establish bounded expectations for users, procurers, and <br>
administrators<br>
<br>
The IO500 follows a two-staged approach. First, there will be a two-week <br>
stabilization period during which we encourage the community to verify <br>
that the benchmark runs properly on a variety of storage systems. During <br>
this period the benchmark may be updated based upon feedback from the <br>
community. The final benchmark will then be released. We expect that <br>
runs compliant with the rules made during the stabilization period will <br>
be valid as a final submission unless a significant defect is found.<br>
<br>
10 Client Node I/O Challenge<br>
The 10 Client Node Challenge is conducted using the regular IO500 <br>
benchmark, however, with the rule that exactly 10 client nodes must be <br>
used to run the benchmark. You may use any shared storage with any <br>
number of servers. We will announce the results in the Production and <br>
Research lists as well as in separate derived lists.<br>
<br>
Birds-of-a-Feather<br>
Once again, we encourage you to submit [2] to join our community, and to <br>
attend the ISC23 BoF [1], where we will announce the new IO500 <br>
Production and Research lists and their 10 client node counterparts. The <br>
current list includes results from twenty different storage system types <br>
and 70 institutions. We hope that the upcoming list grows even more.<br>
<br>
[1] <a href="https://io500.org/pages/bof-isc23">https://io500.org/pages/bof-isc23</a>
<br>
[2] <a href="https://io500.org/submission">https://io500.org/submission</a>
<br>
[3] <a href="https://io500.org/rules-submission">https://io500.org/rules-submission</a>
<br>
<br>
-- <br>
The IO500 Committee<br>
<br>
<br>
<br>
------------------------------<br>
<br>
Subject: Digest Footer<br>
<br>
_______________________________________________<br>
gpfsug-discuss mailing list<br>
gpfsug-discuss at gpfsug.org<br>
<a href="http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org">http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org</a>
<br>
<br>
<br>
------------------------------<br>
<br>
End of gpfsug-discuss Digest, Vol 132, Issue 25<br>
***********************************************<br>
</div>
</span></font></div>
</div>
</body>
</html>