Mailing List Archive

Couple of random questions
Hi all just wondered if anyone can just answer a couple of questions for me.

1 - We are migrating from a number of 7-mode filers to c-mode clusters and the clusters have AFF and FAS HA pairs, management do not want to move someone who is on HDD now onto flash to then move them back to HDD when the disk shelfs are all sorted out. I have suggested using QOS limits while the volumes are on the flash storage but the storage architect has said that it won't work and the customer will see a performance improvement even with QOS on but can't really explain the why behind what he is saying. This has put the whole migration on hold is there any reason a 1000IOPS limit on flash would give better performance that a 1000IOPS limit on HDD the only thing I can see happen is if the requests go over 1000 they will be served up at the start of the second and then there would be a pause until the next second.

2 - Has anyone had been using compaction on HDD aggregates I am thinking we can enable this on our vault volumes but have no idea what the performance hit on the controller would be currently our vault is hanging off of a FAS2554 but this will probably be getting swapped out for a FAS8020 during the working in question 1 so would like to think about saving as much space as is possible.

Regards
Mark
Data Centre Sysadmin Team
Managed Services
Phone:- 02476 694455 Ext 2567
The Sysadmin Team promoting PCMS Values ~Integrity~Respect~Commitment~ ~Continuous Improvement~
The information contained in this e-mail is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. If you are not the intended recipient of this e-mail, the use of this information or any disclosure, copying or distribution is prohibited and may be unlawful. If you received this in error, please contact the sender and delete the material from any computer. The views expressed in this e-mail may not necessarily be the views of the PCMS Group plc and should not be taken as authority to carry out any instruction contained. The PCMS Group reserves the right to monitor and examine the content of all e-mails.

The PCMS Group plc is a company registered in England and Wales with company number 1459419 whose registered office is at PCMS House, Torwood Close, Westwood Business Park, Coventry CV4 8HX, United Kingdom. VAT No: GB 705338743
Re: Couple of random questions [ In reply to ]
There is likely a little truth to #1. Even though you can limit the # of
IOPS, With SSDs the response time per IOP will drop drastically, usually to
sub 1ms timeframes.

That is not always the case with HDDs. The SSDs *will* perform and your
customers may see the improvement!

As for 2, I've only done compaction on SSD aggregates. Turning on some of
the efficiency items for HDD aggregates can put a load on the CPU. If you
are going to use them on HDDs, be judicious. Start with one or two volumes
and see if the performance impact is acceptable.
It seems to be a much longer process to undo an efficiency than it takes to
implement.

Take care

--tmac

*Tim McCarthy, **Principal Consultant*

*Proud Member of the #NetAppATeam <https://twitter.com/NetAppATeam>*

*I Blog at TMACsRack <https://tmacsrack.wordpress.com/>*




On Tue, Jul 4, 2017 at 7:53 AM, Mark Saunders <Mark.Saunders@pcmsgroup.com>
wrote:

> Hi all just wondered if anyone can just answer a couple of questions for
> me.
>
>
>
> 1 – We are migrating from a number of 7-mode filers to c-mode clusters and
> the clusters have AFF and FAS HA pairs, management do not want to move
> someone who is on HDD now onto flash to then move them back to HDD when the
> disk shelfs are all sorted out. I have suggested using QOS limits while the
> volumes are on the flash storage but the storage architect has said that it
> won’t work and the customer will see a performance improvement even with
> QOS on but can’t really explain the why behind what he is saying. This has
> put the whole migration on hold is there any reason a 1000IOPS limit on
> flash would give better performance that a 1000IOPS limit on HDD the only
> thing I can see happen is if the requests go over 1000 they will be served
> up at the start of the second and then there would be a pause until the
> next second.
>
>
>
> 2 – Has anyone had been using compaction on HDD aggregates I am thinking
> we can enable this on our vault volumes but have no idea what the
> performance hit on the controller would be currently our vault is hanging
> off of a FAS2554 but this will probably be getting swapped out for a
> FAS8020 during the working in question 1 so would like to think about
> saving as much space as is possible.
>
>
>
> Regards
>
> Mark
>
> Data Centre Sysadmin Team
>
> Managed Services
>
> Phone:- 02476 694455 Ext 2567
>
> The Sysadmin Team promoting PCMS Values ~Integrity~Respect~Commitment~
> ~Continuous Improvement~
>
> The information contained in this e-mail is intended only for the person
> or entity to which it is addressed and may contain confidential and/or
> privileged material. If you are not the intended recipient of this e-mail,
> the use of this information or any disclosure, copying or distribution is
> prohibited and may be unlawful. If you received this in error, please
> contact the sender and delete the material from any computer. The views
> expressed in this e-mail may not necessarily be the views of the PCMS Group
> plc and should not be taken as authority to carry out any instruction
> contained. The PCMS Group reserves the right to monitor and examine the
> content of all e-mails.
>
>
>
> The PCMS Group plc is a company registered in England and Wales with
> company number 1459419 whose registered office is at PCMS House, Torwood
> Close, Westwood Business Park, Coventry CV4 8HX, United Kingdom. VAT No: GB
> 705338743
>
>
>
> _______________________________________________
> Toasters mailing list
> Toasters@teaparty.net
> http://www.teaparty.net/mailman/listinfo/toasters
>
>
RE: Couple of random questions [ In reply to ]
I would be interested in any results regarding compaction on HDD’s.

Especially with a vault dest as I don’t think you can switch it on for your destination unless it is enabled at source??? (as with dedupe & compression)

From: toasters-bounces@teaparty.net [mailto:toasters-bounces@teaparty.net] On Behalf Of tmac
Sent: 04 July 2017 13:57
To: Mark Saunders
Cc: toasters@teaparty.net
Subject: Re: Couple of random questions

There is likely a little truth to #1. Even though you can limit the # of IOPS, With SSDs the response time per IOP will drop drastically, usually to sub 1ms timeframes.

That is not always the case with HDDs. The SSDs *will* perform and your customers may see the improvement!

As for 2, I've only done compaction on SSD aggregates. Turning on some of the efficiency items for HDD aggregates can put a load on the CPU. If you are going to use them on HDDs, be judicious. Start with one or two volumes and see if the performance impact is acceptable.
It seems to be a much longer process to undo an efficiency than it takes to implement.

Take care

--tmac

Tim McCarthy, Principal Consultant

Proud Member of the #NetAppATeam<https://twitter.com/NetAppATeam>

I Blog at TMACsRack<https://tmacsrack.wordpress.com/>




On Tue, Jul 4, 2017 at 7:53 AM, Mark Saunders <Mark.Saunders@pcmsgroup.com<mailto:Mark.Saunders@pcmsgroup.com>> wrote:
Hi all just wondered if anyone can just answer a couple of questions for me.

1 – We are migrating from a number of 7-mode filers to c-mode clusters and the clusters have AFF and FAS HA pairs, management do not want to move someone who is on HDD now onto flash to then move them back to HDD when the disk shelfs are all sorted out. I have suggested using QOS limits while the volumes are on the flash storage but the storage architect has said that it won’t work and the customer will see a performance improvement even with QOS on but can’t really explain the why behind what he is saying. This has put the whole migration on hold is there any reason a 1000IOPS limit on flash would give better performance that a 1000IOPS limit on HDD the only thing I can see happen is if the requests go over 1000 they will be served up at the start of the second and then there would be a pause until the next second.

2 – Has anyone had been using compaction on HDD aggregates I am thinking we can enable this on our vault volumes but have no idea what the performance hit on the controller would be currently our vault is hanging off of a FAS2554 but this will probably be getting swapped out for a FAS8020 during the working in question 1 so would like to think about saving as much space as is possible.

Regards
Mark
Data Centre Sysadmin Team
Managed Services
Phone:- 02476 694455 Ext 2567
The Sysadmin Team promoting PCMS Values ~Integrity~Respect~Commitment~ ~Continuous Improvement~
The information contained in this e-mail is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. If you are not the intended recipient of this e-mail, the use of this information or any disclosure, copying or distribution is prohibited and may be unlawful. If you received this in error, please contact the sender and delete the material from any computer. The views expressed in this e-mail may not necessarily be the views of the PCMS Group plc and should not be taken as authority to carry out any instruction contained. The PCMS Group reserves the right to monitor and examine the content of all e-mails.

The PCMS Group plc is a company registered in England and Wales with company number 1459419 whose registered office is at PCMS House, Torwood Close, Westwood Business Park, Coventry CV4 8HX, United Kingdom. VAT No: GB 705338743


_______________________________________________
Toasters mailing list
Toasters@teaparty.net<mailto:Toasters@teaparty.net>
http://www.teaparty.net/mailman/listinfo/toasters
RE: Couple of random questions [ In reply to ]
#2 - I have done exactly this.
Our main backup cluster is a 4 node cluster: 8020 and 2240
I enabled compaction on the 8020 aggregates since it had cpu to burn.
Current savings:
NcController Name AggregateUsedSizeGB CompactionSpaceSavedGB
------------ ---- ------------------- ----------------------
******************* aggr1_n01_SATA3000 137428 8499
******************* aggr1_n02_SATA4000 167102 5691

m

From: toasters-bounces@teaparty.net [mailto:toasters-bounces@teaparty.net] On Behalf Of Mark Saunders
Sent: Tuesday, July 4, 2017 6:54 AM
To: toasters@teaparty.net
Subject: Couple of random questions

Hi all just wondered if anyone can just answer a couple of questions for me.

1 - We are migrating from a number of 7-mode filers to c-mode clusters and the clusters have AFF and FAS HA pairs, management do not want to move someone who is on HDD now onto flash to then move them back to HDD when the disk shelfs are all sorted out. I have suggested using QOS limits while the volumes are on the flash storage but the storage architect has said that it won't work and the customer will see a performance improvement even with QOS on but can't really explain the why behind what he is saying. This has put the whole migration on hold is there any reason a 1000IOPS limit on flash would give better performance that a 1000IOPS limit on HDD the only thing I can see happen is if the requests go over 1000 they will be served up at the start of the second and then there would be a pause until the next second.

2 - Has anyone had been using compaction on HDD aggregates I am thinking we can enable this on our vault volumes but have no idea what the performance hit on the controller would be currently our vault is hanging off of a FAS2554 but this will probably be getting swapped out for a FAS8020 during the working in question 1 so would like to think about saving as much space as is possible.

Regards
Mark
Data Centre Sysadmin Team
Managed Services
Phone:- 02476 694455 Ext 2567
The Sysadmin Team promoting PCMS Values ~Integrity~Respect~Commitment~ ~Continuous Improvement~
The information contained in this e-mail is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. If you are not the intended recipient of this e-mail, the use of this information or any disclosure, copying or distribution is prohibited and may be unlawful. If you received this in error, please contact the sender and delete the material from any computer. The views expressed in this e-mail may not necessarily be the views of the PCMS Group plc and should not be taken as authority to carry out any instruction contained. The PCMS Group reserves the right to monitor and examine the content of all e-mails.

The PCMS Group plc is a company registered in England and Wales with company number 1459419 whose registered office is at PCMS House, Torwood Close, Westwood Business Park, Coventry CV4 8HX, United Kingdom. VAT No: GB 705338743