Date: Mon, 08 Nov 2010 09:37:52 +0100 From: "Svein Skogen (Listmail account)" <svein-listmail@stillbilde.net> To: freebsd-questions@freebsd.org Subject: Re: zfs performance issues with iscsi (istgt) Message-ID: <4CD7B6E0.2010501@stillbilde.net> In-Reply-To: <340170.48770.qm@web113309.mail.gq1.yahoo.com> References: <340170.48770.qm@web113309.mail.gq1.yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
This is an OpenPGP/MIME signed message (RFC 2440 and 3156)
--------------enig7E358C8745EC838ABC4580B7
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
On 08.11.2010 09:13, DJ wrote:
>=20
> After scratching my head for a few weeks, I've decided to ask for some =
help.
>=20
> First, I've got two machines connected by gigabit ethernet, network per=
formance is not a problem as I am able to substantially saturate the wire=
when not using iscsi [say iperf] or ftp. Both systems are 8.1-RELENG. Th=
ey are both multi-core, 8G of RAM.=20
>=20
> Symptoms: When doing writes (size relatively independent) from a client=
to a server via iSCSI I seem to be
> hitting a wall between 18-26MB/s of write. This can be repeated contin=
uously whether doing a newfs on a 2TB iscsi volume or doing a dd from /de=
v/zero to the iscsi target. I haven't compared read performance. What ori=
ginally put me on to this was watching the newfs *fly* across the screen,=
and then hang for several seconds, and then *fly* again, and
> then pause.=20
I'll snip down this mail a little bit to ask some control questions
(having recently had quite a wrestle with iSCSI myself).
-Is jumbo frames involved? (and enabled on the initiator, all switches
in between, and the target)
-What's the number of PPS (some switches have PPS issues, which becomes
painfully relevant for small block IO)? (I got rid of most of my
problems when I replaced the Netgear GS724Tv3 switch with a Cisco SG-300)=
-Are you running digests?
-Do you have TSO/TSOv2 enabled at the endpoints?
-Does top -HSC reveal anything?
-Does systat -vmstat 1 reveal anything?
-What's the ICMP (ping) roundtrip times between the initiator and target
IPs?
//Svein
--=20
--------+-------------------+-------------------------------
/"\ |Svein Skogen | svein@d80.iso100.no
\ / |Solberg =D8stli 9 | PGP Key: 0xE5E76831
X |2020 Skedsmokorset | svein@jernhuset.no
/ \ |Norway | PGP Key: 0xCE96CE13
| | svein@stillbilde.net
ascii | | PGP Key: 0x58CD33B6
ribbon |System Admin | svein-listmail@stillbilde.net
Campaign|stillbilde.net | PGP Key: 0x22D494A4
+-------------------+-------------------------------
|msn messenger: | Mobile Phone: +47 907 03 575
|svein@jernhuset.no | RIPE handle: SS16503-RIPE
--------+-------------------+-------------------------------
A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
------------------------------------------------------------
Picture Gallery:
https://gallery.stillbilde.net/v/svein/
------------------------------------------------------------
--------------enig7E358C8745EC838ABC4580B7
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAkzXtuQACgkQODUnwSLUlKSldgCgn74zkXjfsXT0MRfjurLG8PF5
KlgAn3ycNZSiblM9f3lb1qFNj93ckTd0
=U+m9
-----END PGP SIGNATURE-----
--------------enig7E358C8745EC838ABC4580B7--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4CD7B6E0.2010501>