Re: FreeBSD iscsi target
On 0703T1615, Craig Rodrigues wrote: On Tue, Jul 1, 2014 at 2:12 AM, Edward Tomasz Napierała tr...@freebsd.org wrote: In 10-STABLE there is a way to control access based on initiator name and IP address. Edward, Out of curiousity, what kinds of interop testing do you do when you implement the iSCSI code in FreeBSD? As for the target, I wrote a script to test it against both old and new FreeBSD initiators, Linux initiator (Open-iSCSI) and Solaris one; you can find it at tools/regression/iscsi/. I also did manual testing with Windows XP and Windows Vista. I don't remember if I actually succeeded to do any testing with ESX (trying to run ESX under Fusion is not such a good idea, it turns out), but I got a 3rd party report that it worked correctly. As for the initiator, I did manual testing against istgt, LIO (Linux) and COMSTAR (Solaris). I work on FreeNAS at iXsystems, and we have found that iSCSI is a complex protocol, No kidding :-) and there are interop issues, especially with VMWare ESX. Luckily I see that Alexander Motin has been working with you to commit fixes to the iSCSI code, which help. I've rolled an experimental FreeNAS image based on FreeBSD 10 at svn revision r268201 if you want to give it a try: http://download.freenas.org/nightlies/10.0.0/ALPHA/20140703/ For now I'm swamped with work on autofs, but I'd definitely want to redo all the testing before 10.1 - last time I did it was just before 10.0. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Sun, Jul 6, 2014 at 12:08 PM, Edward Tomasz Napierała tr...@freebsd.org wrote: As for the target, I wrote a script to test it against both old and new FreeBSD initiators, Linux initiator (Open-iSCSI) and Solaris one; you can find it at tools/regression/iscsi/. I also did manual testing with Windows XP and Windows Vista. I don't remember if I actually succeeded to do any testing with ESX (trying to run ESX under Fusion is not such a good idea, it turns out), but I got a 3rd party report that it worked correctly. As for the initiator, I did manual testing against istgt, LIO (Linux) and COMSTAR (Solaris). For now I'm swamped with work on autofs, but I'd definitely want to redo all the testing before 10.1 - last time I did it was just before 10.0. Good stuff!When you have more time, I highly recommend that you collaborate with iXsystems (contact d...@ixsystems.com ) , since they can help you with doing interop testing with VMWare ESX and Windows Server environments. It is a lot of work to set up those environments and do the testing, but those environments are what are most heavily used in the real world data centers, and a number of customers are interested in using FreeNAS/TrueNAS (FreeBSD) as a backend storage to those operating systems. -- Craig ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 03, 2014 at 08:39:42PM -0700, Kevin Oberman wrote: In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Yep. It is really crappy LAGG (fixed three-tupple hash... yuck!) and is really nothing but 4 10G Ethernet ports using a 40G PHY in yhe 4x10G form. Note that they don't make any claim of 802.3ba compliance. It only states that 40 Gigabit Ethernet is now part of the IEEE 802.3ba standard. So it is, but this device almost certainly predates the completion of the standard to get a product for which there was great demand. It's a data center product and for typical cases of large numbers of small flow, it should do the trick. Probably does not interoperate with true 80-2.3ba hardware, either. My boss at the time I retired last November was on the committee that wrote 802.3ba. He would be a good authority on whether the standard has any vague wording that would allow this, but he retired 5 month after I did and I have no contact information for him. But I'm pretty sure that there is no way that this is legitimate 40G Ethernet. 802.3ba describe only end point of ethernet. ASIC, internal details of implemetations NICs, switches, fabrics -- out of standart scope. Bottleneck can be in any point of packet flow. In first pappers of netmap test demonstarated NIC can't do saturation of 10G in one stream 64 bytes packet -- need use multiple rings for transmit. I think need use general rule: one flow transfer can hit performance limitation. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Fri, Jul 4, 2014 at 12:16 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Thu, Jul 03, 2014 at 08:39:42PM -0700, Kevin Oberman wrote: In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Yep. It is really crappy LAGG (fixed three-tupple hash... yuck!) and is really nothing but 4 10G Ethernet ports using a 40G PHY in yhe 4x10G form. Note that they don't make any claim of 802.3ba compliance. It only states that 40 Gigabit Ethernet is now part of the IEEE 802.3ba standard. So it is, but this device almost certainly predates the completion of the standard to get a product for which there was great demand. It's a data center product and for typical cases of large numbers of small flow, it should do the trick. Probably does not interoperate with true 80-2.3ba hardware, either. My boss at the time I retired last November was on the committee that wrote 802.3ba. He would be a good authority on whether the standard has any vague wording that would allow this, but he retired 5 month after I did and I have no contact information for him. But I'm pretty sure that there is no way that this is legitimate 40G Ethernet. 802.3ba describe only end point of ethernet. ASIC, internal details of implemetations NICs, switches, fabrics -- out of standart scope. Bottleneck can be in any point of packet flow. In first pappers of netmap test demonstarated NIC can't do saturation of 10G in one stream 64 bytes packet -- need use multiple rings for transmit. that was actually just a configuration issue which since then has been resolved. The 82599 can do 14.88 Mpps on a single ring (and is the only 10G nic i have encountered who can do so). Besides, performance with short packets has nothing to do with the case you were discussing, namely throughput for a single large flow. I think need use general rule: one flow transfer can hit performance limitation. This is neither a useful nor it is restricted to a single flow. Everything can underperform depending on the hw/sw configuration, but not necessarily has to. cheers luigi ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Fri, Jul 04, 2014 at 12:25:35PM +0200, Luigi Rizzo wrote: On Fri, Jul 4, 2014 at 12:16 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Thu, Jul 03, 2014 at 08:39:42PM -0700, Kevin Oberman wrote: In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Yep. It is really crappy LAGG (fixed three-tupple hash... yuck!) and is really nothing but 4 10G Ethernet ports using a 40G PHY in yhe 4x10G form. Note that they don't make any claim of 802.3ba compliance. It only states that 40 Gigabit Ethernet is now part of the IEEE 802.3ba standard. So it is, but this device almost certainly predates the completion of the standard to get a product for which there was great demand. It's a data center product and for typical cases of large numbers of small flow, it should do the trick. Probably does not interoperate with true 80-2.3ba hardware, either. My boss at the time I retired last November was on the committee that wrote 802.3ba. He would be a good authority on whether the standard has any vague wording that would allow this, but he retired 5 month after I did and I have no contact information for him. But I'm pretty sure that there is no way that this is legitimate 40G Ethernet. 802.3ba describe only end point of ethernet. ASIC, internal details of implemetations NICs, switches, fabrics -- out of standart scope. Bottleneck can be in any point of packet flow. In first pappers of netmap test demonstarated NIC can't do saturation of 10G in one stream 64 bytes packet -- need use multiple rings for transmit. ?that was actually just a configuration issue which since then has been ?resolved. The 82599 can do 14.88 Mpps on a single ring (and is the only 10G nic i have encountered who can do so). Thanks for clarification. Besides, performance with short packets has nothing to do with the case you were discussing, namely throughput for a single large flow. This is only illustration about hardware limitation. Perforamnce may be not only bandwidth limited, but interrupt/pps (per flow) limited. I think need use general rule: one flow transfer can hit performance limitation. ?This is neither a useful nor it is restricted to a single flow. Everything can underperform depending on the hw/sw configuration, but not necessarily has to. Yes. And estimate to ideal hw/sw configuration and enviroment -- bad think. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 3, 2014 at 12:06 AM, Kevin Oberman rkober...@gmail.com wrote: On Wed, Jul 2, 2014 at 1:36 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Wed, Jul 02, 2014 at 12:51:59PM -0700, Kevin Oberman wrote: On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. As I know (may be wrong) 40G is bundled 4x10G link. For prevent packet reordering (when run over diferrent link) all packets from one sessoin must be routed to same link. Same issuse for Etherchannel. No, 40G Ethernet is single channel from the interface perspective.. What my be confusing you is that they may use lanes which, for 40G, are 10.3125G. But, unlike the case with Etherchannel, these lanes are hidden from the MAC. The interface deals with a single stream and parcels it out over the 10G (or 25G) lanes. All 100G optical links use multiple lanes (4x25G or 10x10G), but 40G my use either a single 40G lane for distances of up to 2km or 4x10G for longer runs. Since, in most cases, 40G is used within a data center or to connect to wave gear for DWDM transmission over very long distances, most runs are under 2km, so a single 40G lane may be used. When 4 lanes are used, a ribbon cable is required to assure that all optical or copper paths are exactly the same length. Since the PMD is designed to know about and use these lanes for a single channel, the issue of packet re-ordering is not present and the protocol layers above the physical are unaware of how many lanes are used. Wikipedia has a fairly good discussion under the unfortunate title of 100 Gigabit Ethernet https://en.wikipedia.org/wiki/100_Gigabit_Ethernet. Regardless of the title, the article covers both 40 and 100 Gigabit specifications as both were specified on the same standard, 802.3ba. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf --Nikolay ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 03, 2014 at 09:31:45AM +0100, Nikolay Denev wrote: On Thu, Jul 3, 2014 at 12:06 AM, Kevin Oberman rkober...@gmail.com wrote: On Wed, Jul 2, 2014 at 1:36 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Wed, Jul 02, 2014 at 12:51:59PM -0700, Kevin Oberman wrote: On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. As I know (may be wrong) 40G is bundled 4x10G link. For prevent packet reordering (when run over diferrent link) all packets from one sessoin must be routed to same link. Same issuse for Etherchannel. No, 40G Ethernet is single channel from the interface perspective.. What my be confusing you is that they may use lanes which, for 40G, are 10.3125G. But, unlike the case with Etherchannel, these lanes are hidden from the MAC. The interface deals with a single stream and parcels it out over the 10G (or 25G) lanes. All 100G optical links use multiple lanes (4x25G or 10x10G), but 40G my use either a single 40G lane for distances of up to 2km or 4x10G for longer runs. Since, in most cases, 40G is used within a data center or to connect to wave gear for DWDM transmission over very long distances, most runs are under 2km, so a single 40G lane may be used. When 4 lanes are used, a ribbon cable is required to assure that all optical or copper paths are exactly the same length. Since the PMD is designed to know about and use these lanes for a single channel, the issue of packet re-ordering is not present and the protocol layers above the physical are unaware of how many lanes are used. Wikipedia has a fairly good discussion under the unfortunate title of 100 Gigabit Ethernet https://en.wikipedia.org/wiki/100_Gigabit_Ethernet. Regardless of the title, the article covers both 40 and 100 Gigabit specifications as both were specified on the same standard, 802.3ba. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 3, 2014 at 10:13 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Thu, Jul 03, 2014 at 09:31:45AM +0100, Nikolay Denev wrote: On Thu, Jul 3, 2014 at 12:06 AM, Kevin Oberman rkober...@gmail.com wrote: On Wed, Jul 2, 2014 at 1:36 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Wed, Jul 02, 2014 at 12:51:59PM -0700, Kevin Oberman wrote: On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. As I know (may be wrong) 40G is bundled 4x10G link. For prevent packet reordering (when run over diferrent link) all packets from one sessoin must be routed to same link. Same issuse for Etherchannel. No, 40G Ethernet is single channel from the interface perspective.. What my be confusing you is that they may use lanes which, for 40G, are 10.3125G. But, unlike the case with Etherchannel, these lanes are hidden from the MAC. The interface deals with a single stream and parcels it out over the 10G (or 25G) lanes. All 100G optical links use multiple lanes (4x25G or 10x10G), but 40G my use either a single 40G lane for distances of up to 2km or 4x10G for longer runs. Since, in most cases, 40G is used within a data center or to connect to wave gear for DWDM transmission over very long distances, most runs are under 2km, so a single 40G lane may be used. When 4 lanes are used, a ribbon cable is required to assure that all optical or copper paths are exactly the same length. Since the PMD is designed to know about and use these lanes for a single channel, the issue of packet re-ordering is not present and the protocol layers above the physical are unaware of how many lanes are used. Wikipedia has a fairly good discussion under the unfortunate title of 100 Gigabit Ethernet https://en.wikipedia.org/wiki/100_Gigabit_Ethernet. Regardless of the title, the article covers both 40 and 100 Gigabit specifications as both were specified on the same standard, 802.3ba. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Interesting, however this seems like implementation specific detail, and not limitation of native 40Gbit ethernet. Still, it's something that one must be aware of (esp when dealing with Cisco gear :) ) I wonder why they are not doing something like this : http://blog.ipspace.net/2011/04/brocade-vcs-fabric-has-almost-perfect.html --Nikolay ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 03, 2014 at 10:35:55AM +0100, Nikolay Denev wrote: I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Interesting, however this seems like implementation specific detail, and not limitation of native 40Gbit ethernet. I see some perfomance tests on solaris and 40G link. In this test perfomance limited about 10Gbit per flow. May be I found links to this test. May be some NIC's implementation specific detail also limited performance per flow. Still, it's something that one must be aware of (esp when dealing with Cisco gear :) ) I wonder why they are not doing something like this : http://blog.ipspace.net/2011/04/brocade-vcs-fabric-has-almost-perfect.html --Nikolay ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
Which NIC? -a On 3 July 2014 03:29, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Thu, Jul 03, 2014 at 10:35:55AM +0100, Nikolay Denev wrote: I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Interesting, however this seems like implementation specific detail, and not limitation of native 40Gbit ethernet. I see some perfomance tests on solaris and 40G link. In this test perfomance limited about 10Gbit per flow. May be I found links to this test. May be some NIC's implementation specific detail also limited performance per flow. Still, it's something that one must be aware of (esp when dealing with Cisco gear :) ) I wonder why they are not doing something like this : http://blog.ipspace.net/2011/04/brocade-vcs-fabric-has-almost-perfect.html --Nikolay ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 03, 2014 at 10:28:19AM -0700, Adrian Chadd wrote: Which NIC? I am can't find again this forum posts (last time I find -- year ago). May be this http://hardforum.com/showthread.php?t=1662769 In this case -- Mellanox QDR ConnectX2 Infiniband. On 3 July 2014 03:29, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Thu, Jul 03, 2014 at 10:35:55AM +0100, Nikolay Denev wrote: I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Interesting, however this seems like implementation specific detail, and not limitation of native 40Gbit ethernet. I see some perfomance tests on solaris and 40G link. In this test perfomance limited about 10Gbit per flow. May be I found links to this test. May be some NIC's implementation specific detail also limited performance per flow. Still, it's something that one must be aware of (esp when dealing with Cisco gear :) ) I wonder why they are not doing something like this : http://blog.ipspace.net/2011/04/brocade-vcs-fabric-has-almost-perfect.html --Nikolay ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Tue, Jul 1, 2014 at 2:12 AM, Edward Tomasz Napierała tr...@freebsd.org wrote: In 10-STABLE there is a way to control access based on initiator name and IP address. Edward, Out of curiousity, what kinds of interop testing do you do when you implement the iSCSI code in FreeBSD? I work on FreeNAS at iXsystems, and we have found that iSCSI is a complex protocol, and there are interop issues, especially with VMWare ESX. Luckily I see that Alexander Motin has been working with you to commit fixes to the iSCSI code, which help. I've rolled an experimental FreeNAS image based on FreeBSD 10 at svn revision r268201 if you want to give it a try: http://download.freenas.org/nightlies/10.0.0/ALPHA/20140703/ -- Craig ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Thu, Jul 3, 2014 at 2:13 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Thu, Jul 03, 2014 at 09:31:45AM +0100, Nikolay Denev wrote: On Thu, Jul 3, 2014 at 12:06 AM, Kevin Oberman rkober...@gmail.com wrote: On Wed, Jul 2, 2014 at 1:36 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Wed, Jul 02, 2014 at 12:51:59PM -0700, Kevin Oberman wrote: On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. As I know (may be wrong) 40G is bundled 4x10G link. For prevent packet reordering (when run over diferrent link) all packets from one sessoin must be routed to same link. Same issuse for Etherchannel. No, 40G Ethernet is single channel from the interface perspective.. What my be confusing you is that they may use lanes which, for 40G, are 10.3125G. But, unlike the case with Etherchannel, these lanes are hidden from the MAC. The interface deals with a single stream and parcels it out over the 10G (or 25G) lanes. All 100G optical links use multiple lanes (4x25G or 10x10G), but 40G my use either a single 40G lane for distances of up to 2km or 4x10G for longer runs. Since, in most cases, 40G is used within a data center or to connect to wave gear for DWDM transmission over very long distances, most runs are under 2km, so a single 40G lane may be used. When 4 lanes are used, a ribbon cable is required to assure that all optical or copper paths are exactly the same length. Since the PMD is designed to know about and use these lanes for a single channel, the issue of packet re-ordering is not present and the protocol layers above the physical are unaware of how many lanes are used. Wikipedia has a fairly good discussion under the unfortunate title of 100 Gigabit Ethernet https://en.wikipedia.org/wiki/100_Gigabit_Ethernet. Regardless of the title, the article covers both 40 and 100 Gigabit specifications as both were specified on the same standard, 802.3ba. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org I found this white paper useful in understanding how this works : http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-726674.pdf In real world Reality is quite different than it actually is. http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-6500-series-switches/white_paper_c11-696669.html See Packet Path Theory of Operation. Ingress Mode. Yep. It is really crappy LAGG (fixed three-tupple hash... yuck!) and is really nothing but 4 10G Ethernet ports using a 40G PHY in yhe 4x10G form. Note that they don't make any claim of 802.3ba compliance. It only states that 40 Gigabit Ethernet is now part of the IEEE 802.3ba standard. So it is, but this device almost certainly predates the completion of the standard to get a product for which there was great demand. It's a data center product and for typical cases of large numbers of small flow, it should do the trick. Probably does not interoperate with true 80-2.3ba hardware, either. My boss at the time I retired last November was on the committee that wrote 802.3ba. He would be a good authority on whether the standard has any vague wording that would allow this, but he retired 5 month after I did and I have no contact information for him. But I'm pretty sure that there is no way that this is legitimate 40G Ethernet. -- R. Kevin Oberman, Network Engineer, Retired E-mail
Re: FreeBSD iscsi target
On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Wed, Jul 02, 2014 at 03:26:09PM +0400, Slawa Olhovchenkov wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. This is not correct. A 40Gb link does not limit a single transfer to 10G. For example, on FreeBSD all common bandwidth benchmarks reach 40GbE line rate with a single TCP connection at mtu 1500. If a single transfer were limited to 10G you'd need 4 connections to get there. The physical signalling is over four lanes so it's easy to split a 40G link into four separate 10G links. But when running as a 40GbE (this is the usual case) the hardware will combine all the lanes into a single 40G data stream, and you get to use all of the bandwidth. Regards, Navdeep ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. I might also suggest looking at Luigi Rizzo's netmap. It is NOT a drop-in replacement for the TCP stack, but a tool that works with many high-speed Ethernet devices to allow very efficient bulk data transfers. You will see lots of discussion of it on net@. It is available for both FreeBSD and Linux. It has become very popular for this sort of thing, but it does require software customization. Normal network operatipns will continue to use the standard network stack. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Wed, Jul 02, 2014 at 12:51:59PM -0700, Kevin Oberman wrote: On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. As I know (may be wrong) 40G is bundled 4x10G link. For prevent packet reordering (when run over diferrent link) all packets from one sessoin must be routed to same link. Same issuse for Etherchannel. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Wed, Jul 2, 2014 at 1:36 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Wed, Jul 02, 2014 at 12:51:59PM -0700, Kevin Oberman wrote: On Wed, Jul 2, 2014 at 4:26 AM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 10:43:08PM -0700, Kevin Oberman wrote: On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session Yes, this case. As I know, single transfer over 40G link limited by 10G. ??? No, not at all. Getting 40G performance over TCP is not easy, but there is no 10G limitation. As I know (may be wrong) 40G is bundled 4x10G link. For prevent packet reordering (when run over diferrent link) all packets from one sessoin must be routed to same link. Same issuse for Etherchannel. No, 40G Ethernet is single channel from the interface perspective.. What my be confusing you is that they may use lanes which, for 40G, are 10.3125G. But, unlike the case with Etherchannel, these lanes are hidden from the MAC. The interface deals with a single stream and parcels it out over the 10G (or 25G) lanes. All 100G optical links use multiple lanes (4x25G or 10x10G), but 40G my use either a single 40G lane for distances of up to 2km or 4x10G for longer runs. Since, in most cases, 40G is used within a data center or to connect to wave gear for DWDM transmission over very long distances, most runs are under 2km, so a single 40G lane may be used. When 4 lanes are used, a ribbon cable is required to assure that all optical or copper paths are exactly the same length. Since the PMD is designed to know about and use these lanes for a single channel, the issue of packet re-ordering is not present and the protocol layers above the physical are unaware of how many lanes are used. Wikipedia has a fairly good discussion under the unfortunate title of 100 Gigabit Ethernet https://en.wikipedia.org/wiki/100_Gigabit_Ethernet. Regardless of the title, the article covers both 40 and 100 Gigabit specifications as both were specified on the same standard, 802.3ba. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. 4. Dynamic Lun allocation/resize Yes. 5. Target redirection It's in Perforce; I'll try to get it into 11-HEAD shortly. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: FreeBSD iscsi target
On Tue, Jul 1, 2014 at 4:13 PM, Slawa Olhovchenkov s...@zxy.spb.ru wrote: On Tue, Jul 01, 2014 at 11:12:52AM +0200, Edward Tomasz Napierala wrote: Hi. I've replied in private, but just for the record: On 0627T0927, Sreenivasa Honnur wrote: Does freebsd iscsi target supports: 1. ACL (access control lists) In 10-STABLE there is a way to control access based on initiator name and IP address. 2. iSNS No; it's one of the iSCSI features that seem to only be used for marketing purposes :-) 3. Multiple connections per session No; see above. I think this is help for 40G links. I assume that you are looking at transfer of large amounts of data over 40G links. Assuming that tis is the case, yes, multiple connections per session can help you. If you have not done so, you should also look at http://fasterdata.es.net. It is a bit linux-centric these days, but still provides a lot of information on moving data efficiently over fast links. IIt is evolving continually, but did not discuss iSCSI last I knew. ESnet has an Nx100G backbone and many users (most notably the LHC at CERN) who regularly move increasingly large volumes of data, often around the clock, making efficient use of available bandwidth is critical. The network researchers there have put in significant efforts in determining how to best optimize such operations. -- R. Kevin Oberman, Network Engineer, Retired E-mail: rkober...@gmail.com ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
FreeBSD iscsi target
Does freebsd iscsi target supports: 1. ACL (access control lists) 2. iSNS 3. Multiple connections per session 4. Dynamic Lun allocation/resize 5. Target redirection ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org