Btrfs RAID-10 performance

Scott Q. qmail at top-consulting.net
Wed Sep 9 16:51:32 EEST 2020 

The 9361-8i does support passthrough ( JBOD mode ). Make sure you
have the latest firmware.

On Wednesday, 09/09/2020 at 03:55 Miloslav Hůla wrote:


Hi, thank you for your reply. I'll continue inline...

Dne 09.09.2020 v 3:15 John Stoffel napsal(a):
> Miloslav> Hello,
> Miloslav> I sent this into the Linux Kernel Btrfs mailing list and I
got reply:
> Miloslav> "RAID-1 would be preferable"
> Miloslav>
(https://lore.kernel.org/linux-btrfs/7b364356-7041-7d18-bd77-f60e0e2e2112@lechevalier.se/T/).
> Miloslav> May I ask you for the comments as from people around the
Dovecot?
> 
> 
> Miloslav> We are using btrfs RAID-10 (/data, 4.7TB) on a physical
Supermicro
> Miloslav> server with Intel(R) Xeon(R) CPU E5-2620 v4 @ 2.10GHz and
125GB of RAM.
> Miloslav> We run 'btrfs scrub start -B -d /data' every Sunday as a
cron task. It
> Miloslav> takes about 50 minutes to finish.
> 
> Miloslav> # uname -a
> Miloslav> Linux imap 4.9.0-12-amd64 #1 SMP Debian 4.9.210-1
(2020-01-20) x86_64
> Miloslav> GNU/Linux
> 
> Miloslav> RAID is a composition of 16 harddrives. Harddrives are
connected via
> Miloslav> AVAGO MegaRAID SAS 9361-8i as a RAID-0 devices. All
harddrives are SAS
> Miloslav> 2.5" 15k drives.
> 
> Can you post the output of "cat /proc/mdstat" or since you say
you're
> using btrfs, are you using their own RAID0 setup?  If so, please
post
> the output of 'btrfs stats' or whatever the command is you use to
view
> layout info?

There is a one PCIe RAID controller in a chasis. AVAGO MegaRAID SAS 
9361-8i. And 16x SAS 15k drives conneced to it. Because the controller

does not support pass-through for the drives, we use 16x RAID-0 on 
controller. So, we get /dev/sda ... /dev/sdp (roughly) in OS. And over

that we have single btrfs RAID-10, composed of 16 devices, mounted as
/data.

We have chosen this wiring for severeal reasons:
- easy to increase a capacity
- easy to replace drives by larger ones
- due to checksuming, btrfs does not need fsck in case of power
failure
- btrfs scrub discovers failing drive sooner than S.M.A.R.T. or RAID 
controller


> Miloslav> Server serves as a IMAP with Dovecot 2.2.27-3+deb9u6, 4104
accounts,
> Miloslav> Mailbox format, LMTP delivery.
> 
> How ofter are these accounts hitting the server?

IMAP serves for a univesity. So there are typical rush hours from 7AM
to 
3PM. Lowers during the evening, almost not used during the night.


> Miloslav> We run 'rsync' to remote NAS daily. It takes about 6.5
hours to finish,
> Miloslav> 12'265'387 files last night.
> 
> That's.... sucky.  So basically you're hitting the drives hard
with
> random IOPs and you're probably running out of performance.  How
much
> space are you using on the filesystem?

It's not so sucky how it seems. rsync runs during the night. And even 
reading is high, server load stays low. We have problems with writes.


> And why not use brtfs send to ship off snapshots instead of using
> rsync?  I'm sure that would be an improvement...

We run backup to external NAS (NetApp) for a disaster recovery
scenario. 
Moreover NAS is spreaded across multiple locations. Then we create NAS

snapshot, tens days backward. All snapshots easily available via NFS 
mount. And NAS capacity is cheaper.


> Miloslav> Last half year, we encoutered into performace
> Miloslav> troubles. Server load grows up to 30 in rush hours, due to
> Miloslav> IO waits. We tried to attach next harddrives (the 838G
ones
> Miloslav> in a list below) and increase a free space by rebalace. I
> Miloslav> think, it helped a little bit, not not so rapidly.
> 
> If you're IOPs bound, but not space bound, then you *really* want to
> get an SSD in there for the indexes and such.  Basically the stuff
> that gets written/read from all the time no matter what, but which
> isn't large in terms of space.

Yes. We are now on 66% capacity. Adding SSD for indexes is our next
step.


> Also, adding in another controller card or two would also probably
> help spread the load across more PCI channels, and reduce contention
> on the SATA/SAS bus as well.

Probably we will wait how SSD helps first, but as you wrote, it is 
possible next step.

> Miloslav> Is this a reasonable setup and use case for btrfs RAID-10?
> Miloslav> If so, are there some recommendations to achieve better
> Miloslav> performance?
> 
> 1. move HOT data to SSD based volume RAID 1 pair.  On a seperate
>     controller.

OK

> 2. add more controllers, which also means you're more redundant in
>     case one controller fails.

OK

> 3. Clone the system and put Dovecot IMAP director in from of the
>     setup.

I still hope that one server can handle 4105 accounts.

> 4. Stop using rsync for copying to your DR site, use the btrfs snap
>     send, or whatever the commands are.

I hope it is not needed in our scenario.

> 5. check which dovecot backend you're using and think about moving
to
>     one which doesn't involve nearly as many files.

Maildir is comfortable for us. Time to time, users call us with: "I 
accidentally deleted the folder" and it is super easy to copy it back 
from backup.

> 6. Find out who your biggest users are, in terms of emails and move
>     them to SSDs if step 1 is too hard to do at first.

OK

> Can you also grab some 'iostat -dhm 30 60'  output, which is 30
> minutes of data over 30 second intervals?  That should help you
narrow
> down which (if any) disk is your hotspot.

OK, thanks for the tip.

> It's not clear to me if you have one big btrfs filesystem, or a
bunch
> of smaller ones stiched together.  In any case, it should be very
easy
> to get better performance here.

I hope I've made it clear above.

> I think someone else mentioned that you should look at your dovecot
> backend, and you should move to the fastest one you can find.
> 
> Good luck!
> John

Thank you for your time and advices!

Kind regards
Milo


> Miloslav> # megaclisas-status
> Miloslav> -- Controller information --
> Miloslav> -- ID | H/W Model                  |
RAM    | Temp | BBU    | Firmware
> Miloslav> c0    | AVAGO MegaRAID SAS 9361-8i | 1024MB | 72C  |
Good   | FW:
> Miloslav> 24.16.0-0082
> 
> Miloslav> -- Array information --
> Miloslav> -- ID | Type   |    Size |  Strpsz | Flags |
DskCache |   Status |  OS
> Miloslav> Path | CacheCade |InProgress
> Miloslav> c0u0  | RAID-0 |    838G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdq | None      |None
> Miloslav> c0u1  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sda | None      |None
> Miloslav> c0u2  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdb | None      |None
> Miloslav> c0u3  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdc | None      |None
> Miloslav> c0u4  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdd | None      |None
> Miloslav> c0u5  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sde | None      |None
> Miloslav> c0u6  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdf | None      |None
> Miloslav> c0u7  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdg | None      |None
> Miloslav> c0u8  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdh | None      |None
> Miloslav> c0u9  | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdi | None      |None
> Miloslav> c0u10 | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdj | None      |None
> Miloslav> c0u11 | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdk | None      |None
> Miloslav> c0u12 | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdl | None      |None
> Miloslav> c0u13 | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdm | None      |None
> Miloslav> c0u14 | RAID-0 |    558G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdn | None      |None
> Miloslav> c0u15 | RAID-0 |    838G |  256 KB | RA,WB
|  Enabled |  Optimal |
> Miloslav> /dev/sdr | None      |None
> 
> Miloslav> -- Disk information --
> Miloslav> -- ID   | Type | Drive
Model                       | Size     |
Status
> Miloslav>       | Speed    | Temp | Slot ID  | LSI ID
> Miloslav> c0u0p0  | HDD  | SEAGATE ST900MP0006 N003WAG0Q3S3  |
837.8 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 53C  | [8:14]   | 32
> Miloslav> c0u1p0  | HDD  | HGST HUC156060CSS200 A3800XV250TJ |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 38C  | [8:0]    | 12
> Miloslav> c0u2p0  | HDD  | HGST HUC156060CSS200 A3800XV3XT4J |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 43C  | [8:1]    | 11
> Miloslav> c0u3p0  | HDD  | HGST HUC156060CSS200 ADB05ZG4XLZU |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 46C  | [8:2]    | 25
> Miloslav> c0u4p0  | HDD  | HGST HUC156060CSS200 A3800XV3DWRL |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 48C  | [8:3]    | 14
> Miloslav> c0u5p0  | HDD  | HGST HUC156060CSS200 A3800XV3XZTL |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 52C  | [8:4]    | 18
> Miloslav> c0u6p0  | HDD  | HGST HUC156060CSS200 A3800XV3VSKJ |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 55C  | [8:5]    | 15
> Miloslav> c0u7p0  | HDD  | SEAGATE ST600MP0006 N003WAF1LWKE  |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 56C  | [8:6]    | 28
> Miloslav> c0u8p0  | HDD  | HGST HUC156060CSS200 A3800XV3XTDJ |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 55C  | [8:7]    | 20
> Miloslav> c0u9p0  | HDD  | HGST HUC156060CSS200 A3800XV3T8XL |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 57C  | [8:8]    | 19
> Miloslav> c0u10p0 | HDD  | HGST HUC156060CSS200 A7030XHL0ZYP |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 61C  | [8:9]    | 23
> Miloslav> c0u11p0 | HDD  | HGST HUC156060CSS200 ADB05ZG4VR3P |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 60C  | [8:10]   | 24
> Miloslav> c0u12p0 | HDD  | SEAGATE ST600MP0006 N003WAF195KA  |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 60C  | [8:11]   | 29
> Miloslav> c0u13p0 | HDD  | SEAGATE ST600MP0006 N003WAF1LTZW  |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 56C  | [8:12]   | 26
> Miloslav> c0u14p0 | HDD  | SEAGATE ST600MP0006 N003WAF1LWH6  |
558.4 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 55C  | [8:13]   | 27
> Miloslav> c0u15p0 | HDD  | SEAGATE ST900MP0006 N003WAG0Q414  |
837.8 Gb | Online,
> Miloslav> Spun Up | 12.0Gb/s | 47C  | [8:15]   | 33
> 
> 
> 
> Miloslav> # btrfs --version
> Miloslav> btrfs-progs v4.7.3
> 
> 
> 
> Miloslav> # btrfs fi show
> Miloslav> Label: 'DATA'  uuid:
5b285a46-e55d-4191-924f-0884fa06edd8
> Miloslav>          Total devices 16 FS bytes used 3.49TiB
> Miloslav>          devid    1 size 558.41GiB used
448.66GiB path /dev/sda
> Miloslav>          devid    2 size 558.41GiB used
448.66GiB path /dev/sdb
> Miloslav>          devid    4 size 558.41GiB used
448.66GiB path /dev/sdd
> Miloslav>          devid    5 size 558.41GiB used
448.66GiB path /dev/sde
> Miloslav>          devid    7 size 558.41GiB used
448.66GiB path /dev/sdg
> Miloslav>          devid    8 size 558.41GiB used
448.66GiB path /dev/sdh
> Miloslav>          devid    9 size 558.41GiB used
448.66GiB path /dev/sdf
> Miloslav>          devid   10 size 558.41GiB used
448.66GiB path /dev/sdi
> Miloslav>          devid   11 size 558.41GiB used
448.66GiB path /dev/sdj
> Miloslav>          devid   13 size 558.41GiB used
448.66GiB path /dev/sdk
> Miloslav>          devid   14 size 558.41GiB used
448.66GiB path /dev/sdc
> Miloslav>          devid   15 size 558.41GiB used
448.66GiB path /dev/sdl
> Miloslav>          devid   16 size 558.41GiB used
448.66GiB path /dev/sdm
> Miloslav>          devid   17 size 558.41GiB used
448.66GiB path /dev/sdn
> Miloslav>          devid   18 size 837.84GiB used
448.66GiB path /dev/sdr
> Miloslav>          devid   19 size 837.84GiB used
448.66GiB path /dev/sdq
> 
> 
> 
> Miloslav> # btrfs fi df /data/
> Miloslav> Data, RAID10: total=3.48TiB, used=3.47TiB
> Miloslav> System, RAID10: total=256.00MiB, used=320.00KiB
> Miloslav> Metadata, RAID10: total=21.00GiB, used=18.17GiB
> Miloslav> GlobalReserve, single: total=512.00MiB, used=0.00B
> 
> 
> 
> Miloslav> I do not attach whole dmesg log. It is almost empty,
without errors.
> Miloslav> Only lines about BTRFS are about relocations, like:
> 
> Miloslav> BTRFS info (device sda): relocating block group
29435663220736 flags 65
> Miloslav> BTRFS info (device sda): found 54460 extents
> Miloslav> BTRFS info (device sda): found 54459 extents
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://dovecot.org/pipermail/dovecot/attachments/20200909/71ffa83f/attachment-0001.html>

More information about the dovecot mailing list