Re: Two node MetroCluster performance issues?

Wouter Vervloesem <[email protected]> Wed, 21 Dec 2022 16:20:20 +0000
Newsgroups gmane.comp.hardware.netapp
Message-ID <[email protected]>
--===============5501462085056278274==
Content-Language: nl-NL
Content-type: multipart/signed;
	protocol="application/pkcs7-signature";
	micalg=sha256;
	boundary="B_3754488017_2890159827"

--B_3754488017_2890159827
Content-type: multipart/alternative;
	boundary="B_3754488017_3089203812"


--B_3754488017_3089203812
Content-type: text/plain;
	charset="UTF-8"
Content-transfer-encoding: quoted-printable

Hi

=20

Your ESXi hosts will only see the paths to the cluster that owns the LUN. B=
ecause you have a two node MetroCluster, only a single node will present the=
 LUN. The DR-partner in the MetroCluster setup doesn=E2=80=99t announce the paths =
until going into switchover mode.=20

=20

So you have a single node on each site, connected with a single connection =
to each switch?

If your ESXi hosts are also connected with a single link to each switch I w=
ouldn=E2=80=99t expect 4 paths per LUN.

=20

=20

The write operation for the NetApp is the same if you do the test for remot=
e or local datastores. The moment the write reaches the NetApp, each write h=
as to be committed by both systems before sending the acknowledgement. The o=
nly difference is the frontend fabric. If you see that order of difference b=
etween remote of local connectivity, I would look into the ISL of your front=
end fabric.

=20

=20

=20

Met vriendelijke groeten,

=20

Wouter Vervloesem

Senior Consultant

=20

Neoria NV

Prins Boudewijnlaan 41 - 2650 Edegem

T +32 3 451 23 82 | M +32 496 52 93 61

=20

Van: Toasters <[email protected]> namens Heino Walther <hw@bear=
dmann.dk>
Datum: dinsdag 13 december 2022 om 22:31
Aan: "[email protected]" <[email protected]>
Onderwerp: Two node MetroCluster performance issues?

=20

Hi there

=20

I have a setup with a two node Fabric MetroCluster with A300 nodes and four=
 Brocade 6510 switches places about 10KM from each other.

Our ESXi 7.0 hosts connects via 16G FC using another four frontend Brocade =
6510 switches.

Our ESXi hosts can see four paths for each LUN they are presented from the =
SVMs.

ESXi show all four paths as Active (I/O) which I find a bit odd, because tw=
o of the paths are remote to the ESXi hosts=E2=80=A6

Both the ESXi and the igroup on the NetApp is configured for ALUA=E2=80=A6 but I =
have googled that there can be issues with this, and maybe we have those iss=
ues=E2=80=A6

=20

The main issue is that if we do a few performance tests across different da=
tastores (local and remote to the ESXi host), we see OK performance for the =
local datastores (800+MB/sec.) but if we try to test against a datastore tha=
t is remote to the ESXi host we see 50-60MB/sec. which is a huge difference =
and leads us to question the setup=E2=80=A6

=20

We are aware that especially writing to a remote datastore will involve a t=
ransfer to the remote controller (10KM) this controller then has to write th=
is to it=E2=80=99s peer (remote) controller (10KM) the remote peer than has to sen=
d an ack. back to the peer (10KM) which then sends an ack. to the ESXi host =
(10KM) so all in all 40KM plus waiting for the systems=E2=80=A6  it is not ideal, =
but is this kind of performance normal?

The two nodes does have an ethernet based cluster interconnect which is cur=
rently linked at 1Gb, and I am beginning to suspect that the data between th=
e two nodes is going via this link?  But the more I think about it, the more=
 it does not make sense?

=20

Our FC interswitch links are running 8Gb, but on the switches we see nothin=
g near saturation of any ports=E2=80=A6 and we of cause also checked for port erro=
rs of any kind=E2=80=A6=20

=20

If anyone has a similar setup, any help would be great=E2=80=A6 we are doing a fe=
w more tests, but we are close to opening a case with NetApp=E2=80=A6

=20

/B


--B_3754488017_3089203812
Content-type: text/html;
	charset="UTF-8"
Content-transfer-encoding: quoted-printable

<html xmlns:o=3D"urn:schemas-microsoft-com:office:office" xmlns:w=3D"urn:schema=
s-microsoft-com:office:word" xmlns:m=3D"http://schemas.microsoft.com/office/20=
04/12/omml" xmlns=3D"http://www.w3.org/TR/REC-html40"><head><meta http-equiv=3DC=
ontent-Type content=3D"text/html; charset=3Dutf-8"><meta name=3DGenerator content=3D=
"Microsoft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
	{font-family:"Century Gothic";
	panose-1:2 11 5 2 2 2 2 2 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;
	mso-ligatures:standardcontextual;
	mso-fareast-language:EN-US;}
span.E-mailStijl19
	{mso-style-type:personal-reply;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-size:10.0pt;}
@page WordSection1
	{size:612.0pt 792.0pt;
	margin:3.0cm 2.0cm 3.0cm 2.0cm;}
div.WordSection1
	{page:WordSection1;}
--></style></head><body lang=3DNL-BE link=3D"#0563C1" vlink=3D"#954F72" style=3D'wo=
rd-wrap:break-word'><div class=3DWordSection1><p class=3DMsoNormal>Hi<o:p></o:p>=
</p><p class=3DMsoNormal><o:p>&nbsp;</o:p></p><p class=3DMsoNormal><span lang=3DEN=
-US>Your ESXi hosts will only see the paths to the cluster that owns the LUN=
. Because you have a two node MetroCluster, only a single node will present =
the LUN. The DR-partner in the MetroCluster setup doesn=E2=80=99t announce the pat=
hs until going into switchover mode. <o:p></o:p></span></p><p class=3DMsoNorma=
l><span lang=3DEN-US><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span lang=
=3DEN-US>So you have a single node on each site, connected with a single conne=
ction to each switch?<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-=
US>If your ESXi hosts are also connected with a single link to each switch I=
 wouldn=E2=80=99t expect 4 paths per LUN.<o:p></o:p></span></p><p class=3DMsoNormal>=
<span lang=3DEN-US><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span lang=3DE=
N-US><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>The wri=
te operation for the NetApp is the same if you do the test for remote or loc=
al datastores. The moment the write reaches the NetApp, each write has to be=
 committed by both systems before sending the acknowledgement. The only diff=
erence is the frontend fabric. If you see that order of difference between r=
emote of local connectivity, I would look into the ISL of your frontend fabr=
ic.<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US><o:p>&nbsp;</o:=
p></span></p><div><p class=3DMsoNormal><span lang=3DEN-US style=3D'mso-ligatures:n=
one;mso-fareast-language:NL'><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal>=
<span lang=3DEN-US style=3D'mso-ligatures:none;mso-fareast-language:NL'><o:p>&nb=
sp;</o:p></span></p><p class=3DMsoNormal><span style=3D'font-size:9.0pt;font-fam=
ily:"Century Gothic",sans-serif;color:#5B5B59;mso-ligatures:none;mso-fareast=
-language:NL'>Met vriendelijke groeten,<o:p></o:p></span></p><p class=3DMsoNor=
mal><span style=3D'font-size:9.0pt;font-family:"Century Gothic",sans-serif;col=
or:#5B5B59;mso-ligatures:none;mso-fareast-language:NL'><o:p>&nbsp;</o:p></sp=
an></p><p class=3DMsoNormal style=3D'background:white'><b><span style=3D'font-fami=
ly:"Century Gothic",sans-serif;color:#1B4371;mso-ligatures:none;mso-fareast-=
language:EN-GB'><a href=3D"https://be.linkedin.com/in/woutervervloesem"><span =
style=3D'color:#1B4371;text-decoration:none'>Wouter Vervloesem</span></a><o:p>=
</o:p></span></b></p><p class=3DMsoNormal style=3D'background:white'><span style=
=3D'font-size:9.0pt;font-family:"Century Gothic",sans-serif;color:#5B5B59;mso-=
ligatures:none;mso-fareast-language:EN-GB'>Senior Consultant<o:p></o:p></spa=
n></p><p class=3DMsoNormal style=3D'background:white'><span style=3D'font-size:9.0=
pt;font-family:"Century Gothic",sans-serif;color:#5B5B59;mso-ligatures:none;=
mso-fareast-language:EN-GB'>&nbsp;<o:p></o:p></span></p><p class=3DMsoNormal s=
tyle=3D'background:white'><b><span style=3D'font-size:10.5pt;color:#5B5B59;mso-l=
igatures:none;mso-fareast-language:NL'><a href=3D"https://www.neoria.be/" targ=
et=3D"_blank" title=3D"Original URL: http://www.neoria.be/. Click or tap if you =
trust this link."><span style=3D'color:#5B5B59;mso-fareast-language:EN-GB;text=
-decoration:none'>Neoria NV</span></a></span></b><b><span style=3D'font-size:1=
0.5pt;color:#5B5B59;mso-ligatures:none;mso-fareast-language:EN-GB'><o:p></o:=
p></span></b></p><p class=3DMsoNormal style=3D'background:white'><span style=3D'fo=
nt-size:9.0pt;font-family:"Century Gothic",sans-serif;color:#5B5B59;mso-liga=
tures:none;mso-fareast-language:EN-GB'>Prins Boudewijnlaan 41&nbsp;-&nbsp;26=
50 Edegem<o:p></o:p></span></p></div><p class=3DMsoNormal><b><span style=3D'font=
-size:9.0pt;font-family:"Century Gothic",sans-serif;color:#5B5B59;mso-ligatu=
res:none'>T</span></b><span style=3D'font-size:9.0pt;font-family:"Century Goth=
ic",sans-serif;color:#5B5B59;mso-ligatures:none'>&nbsp;+32 3 451 23 82&nbsp;=
|&nbsp;<b>M</b>&nbsp;+32 496 52 93 61</span><o:p></o:p></p><p class=3DMsoNorma=
l><o:p>&nbsp;</o:p></p><div style=3D'border:none;border-top:solid #B5C4DF 1.0p=
t;padding:3.0pt 0cm 0cm 0cm'><p class=3DMsoNormal><b><span lang=3DDA style=3D'font=
-size:12.0pt;color:black;mso-ligatures:none;mso-fareast-language:NL'>Van: </=
span></b><span lang=3DDA style=3D'font-size:12.0pt;color:black;mso-ligatures:non=
e;mso-fareast-language:NL'>Toasters &lt;[email protected]&gt; na=
mens Heino Walther &lt;[email protected]&gt;<br><b>Datum: </b>dinsdag 13 decem=
ber 2022 om 22:31<br><b>Aan: </b>&quot;[email protected]&quot; &lt;toast=
[email protected]&gt;<br><b>Onderwerp: </b>Two node MetroCluster performance =
issues?<o:p></o:p></span></p></div><div><p class=3DMsoNormal><span lang=3DDA sty=
le=3D'mso-ligatures:none;mso-fareast-language:NL'><o:p>&nbsp;</o:p></span></p>=
</div><p class=3DMsoNormal><span lang=3DEN-US>Hi there<o:p></o:p></span></p><p c=
lass=3DMsoNormal><span lang=3DEN-US><o:p>&nbsp;</o:p></span></p><p class=3DMsoNorm=
al><span lang=3DEN-US>I have a setup with a two node Fabric MetroCluster with =
A300 nodes and four Brocade 6510 switches places about 10KM from each other.=
<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>Our ESXi 7.0 hosts=
 connects via 16G FC using another four frontend Brocade 6510 switches.<o:p>=
</o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>Our ESXi hosts can see =
four paths for each LUN they are presented from the SVMs.<o:p></o:p></span><=
/p><p class=3DMsoNormal><span lang=3DEN-US>ESXi show all four paths as Active (I=
/O) which I find a bit odd, because two of the paths are remote to the ESXi =
hosts=E2=80=A6<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>Both the E=
SXi and the igroup on the NetApp is configured for ALUA=E2=80=A6 but I have google=
d that there can be issues with this, and maybe we have those issues=E2=80=A6<o:p>=
</o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US><o:p>&nbsp;</o:p></span=
></p><p class=3DMsoNormal><span lang=3DEN-US>The main issue is that if we do a f=
ew performance tests across different datastores (local and remote to the ES=
Xi host), we see OK performance for the local datastores (800+MB/sec.) but i=
f we try to test against a datastore that is remote to the ESXi host we see =
50-60MB/sec. which is a huge difference and leads us to question the setup=E2=80=
=A6<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US><o:p>&nbsp;</o:p>=
</span></p><p class=3DMsoNormal><span lang=3DEN-US>We are aware that especially =
writing to a remote datastore will involve a transfer to the remote controll=
er (10KM) this controller then has to write this to it=E2=80=99s peer (remote) con=
troller (10KM) the remote peer than has to send an ack. back to the peer (10=
KM) which then sends an ack. to the ESXi host (10KM) so all in all 40KM plus=
 waiting for the systems=E2=80=A6&nbsp; it is not ideal, but is this kind of perfo=
rmance normal?<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>The =
two nodes does have an ethernet based cluster interconnect which is currentl=
y linked at 1Gb, and I am beginning to suspect that the data between the two=
 nodes is going via this link?&nbsp; But the more I think about it, the more=
 it does not make sense?<o:p></o:p></span></p><p class=3DMsoNormal><span lang=3D=
EN-US><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>Our FC=
 interswitch links are running 8Gb, but on the switches we see nothing near =
saturation of any ports=E2=80=A6 and we of cause also checked for port errors of a=
ny kind=E2=80=A6 <o:p></o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US><o:p>&n=
bsp;</o:p></span></p><p class=3DMsoNormal><span lang=3DEN-US>If anyone has a sim=
ilar setup, any help would be great=E2=80=A6 we are doing a few more tests, but we=
 are close to opening a case with NetApp=E2=80=A6<o:p></o:p></span></p><p class=3DMs=
oNormal><span lang=3DEN-US><o:p>&nbsp;</o:p></span></p><p class=3DMsoNormal><spa=
n lang=3DEN-US>/B<o:p></o:p></span></p></div></body></html>

--B_3754488017_3089203812--

--B_3754488017_2890159827
Content-type: application/pkcs7-signature; name="smime.p7s"
Content-transfer-encoding: base64
Content-disposition: attachment;
	filename="smime.p7s"
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--B_3754488017_2890159827--

--===============5501462085056278274==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Toasters mailing list
[email protected]
https://www.teaparty.net/mailman/listinfo/toasters
--===============5501462085056278274==--