[lustre-discuss] [EXTERNAL] Re: Write Performance is Abnormal for max_dirty_mb Value of 2047

[Andreas Dilger]

The filesystem looks like a test system (only 200GB OSTs, the whole filesystem is smaller than a single HDD), so it doesn't make sense to be testing Lustre 2.9 at this point.  At a very minimum it should be 2.12.8, but for a new test filesystem it makes sense to try out master (2.15.0-RC2) since it will be the new LTS release very soon.

Cheers, Andreas.

On Mar 29, 2022, at 11:08, Patrick Farrell via lustre-discuss <lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org>> wrote:

Hasan,

That's quite old - 2.9 was released nearly 6 years ago and wasn't a maintenance release, so had no updates after release.  I expect some of those bugs are present there.  A newer version should help.

-Patrick
________________________________
From: Hasan Rashid <mrashid2 at uncc.edu<mailto:mrashid2 at uncc.edu>>
Sent: Tuesday, March 29, 2022 11:50 AM
To: Patrick Farrell <pfarrell at ddn.com<mailto:pfarrell at ddn.com>>
Cc: lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org> <lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org>>
Subject: Re: [EXTERNAL] Re: [lustre-discuss] Write Performance is Abnormal for max_dirty_mb Value of 2047

Hi Patrick,

Thank you so much for getting back to me. The Lustre installation was 2.9.0. I am working to run the workloads on Lustre 2.12 release.

I have 1 machine for MGS and MDS and 4 machines for OSS. In each OSS, I have 2 OSTs and each of them is 200 GiB in size. All the OSTs are mounted on SSD. The striping layout is the default one. Since for the workload, we have one large file of 5 GiB in size, the whole resides in a single OST. We also verified that by looking at the rpc_stats.

<image.png>

Following is the hardware configuration for all the machines.

<image.png>

Please let me know if you need further information.

Thanks,
Hasan

On Sun, Mar 27, 2022 at 1:54 PM Patrick Farrell <pfarrell at ddn.com<mailto:pfarrell at ddn.com>> wrote:
[Caution: Email from External Sender. Do not click or open links or attachments unless you know this sender.]

Hasan,

Historically, there have been several bugs related to write grant when max_dirty_mb is set to large values (depending on a few other details of system setup).

Write grant allows the client to write data in to memory and write it out asynchronously.  When write grant is not available to the client, the client is forced to do sync writes at small sizes.  The result looks exactly like this, write performance drops severely.

Depending on what version you're running, you may not have fixes for these bugs.  You could either try a newer Lustre version (you didn't mention what you're running) or just use a smaller value of max_dirty_mb.

I am surprised to see you're still seeing a speedup from max_dirty_mb values over 1 GiB in size.

Can you describe your system a bit more?  How many OSTs do you have and how many stripes are you using?  max_dirty_mb is a per OST value on the client, not a global one.

-Patrick
________________________________
From: lustre-discuss <lustre-discuss-bounces at lists.lustre.org<mailto:lustre-discuss-bounces at lists.lustre.org>> on behalf of Hasan Rashid via lustre-discuss <lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org>>
Sent: Friday, March 25, 2022 11:45 AM
To: lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org> <lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org>>
Subject: [lustre-discuss] Write Performance is Abnormal for max_dirty_mb Value of 2047

Hi Everyone,

As the manual suggests, the valid value range for max_dirty_mb is the values larger than 0 and smaller than the lesser of 2048 MiB or 1/4 of client RAM. In my system, the client's RAM is 196 GiB. So, the maximum valid value for max_dirty_mb(mdm) is 2047 MiB.

However, when we set the max_dirty_mb value to 2047, we see very low write throughput for multiple Filebench workloads that we have tested so far. I am providing details for one example of the tested workload below.

Workload Detail: We are doing only random write operation of 1MiB size from one process and one thread to a single large file of 5GiB size.

Observed Result: As you can see from the below diagram, as we increase the mdm value from 768 to 1792 by an amount of 256 in each step, the write throughput has increased gradually. However, for the mdm value of 2047, the result dropped very significantly. The observation holds true for all the workloads we tested so far.


[https://lh3.googleusercontent.com/iEqpGNZhI9r9jJCLq0rWPvFADJRXkKKKZnyCV_8m3nhiHggNqWU9d_7WTUU0yeb011nxjULF4_iLkI7TIc0qe5el11PJI3i9Jot9KveXUil98A_UEnBojFqAHfK94ve1foQT39m2]

I am unable to figure out why we would have such low performance at the mdm value of 2047. Please share any insights you have that would be helpful for me to understand the aforementioned scenario.

Best Wishes,
Md Hasanur Rashid
_______________________________________________
lustre-discuss mailing list
lustre-discuss at lists.lustre.org<mailto:lustre-discuss at lists.lustre.org>
http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org

Cheers, Andreas
--
Andreas Dilger
Lustre Principal Architect
Whamcloud







-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20220330/7b362a85/attachment-0001.html>