So following my recent blog post on SANSlide WAN optimization appliances for use with Storwize replication, IBM has just announced Storwize 7.2 (available December) which includes not only replication natively over IP networks (licensed as Global Mirror/Metro Mirror) but also has SANslide WAN optimization built-in for free. i.e. to get the benefits of WAN optimization you no longer need to purchase Riverbed or Cisco WAAS or SANSlide appliances.
Admittedly, Global Mirror was a little behind the times in getting to a native IP implementation, but having got there, the developers obviously decided they wanted to do it in style and take the lead in this space, by offering a more optimized WAN replication experience than any of our competitors.
The industry problem with TCP/IP latency is the time it takes to acknowledge that your packets have arrived at the other end. You can’t send the next set of packets until you get that acknowledgement back. So on a high latency network you end up spending a lot of your time waiting, which means you can’t take proper advantage of the available bandwidth. Effective bandwidth usage can sometimes be reduced to only 20% of the actual bandwidth you are paying for.
The first time I heard this story was actually back in the mid-90’s from a telco network engineer. His presentation was entitled something like “How latency can steal your bandwidth”.
SANSlide mitigates latency by virtualising the pipe with many connections. While one connection is waiting for the ACK another is sending data. Using many connections, the pipe can often be filled more than 95%.
If you have existing FCIP routers you don’t need to rush out and switch over to IP replication with SANSlide, especially if your latency is reasonably low, but if you do have a high latency network it would be worth discussing your options with your local IBM Storwize expert. It might depend on the sophistication of your installed FCIP routers. Brocade for example suggests that the IBM SAN06B-R is pretty good at WAN optimization. So the graph below does not necessarily apply to all FCIP routers.
When you next compare long distance IBM Storwize replication to our competitors’ offerings, you might want to ask them to include the cost of WAN optimization appliances to get a full apples for apples comparison, or you might want to take into account that with IBM Storwize you will probably need a lot less bandwidth to achieve the same RPO.
Even when others do include products like Riverbed appliances with their offerings, SANSlide still has the advantage of being completely data-agnostic, so it doesn’t get confused or slow down when transmitting encrypted or compressed data like most other WAN optimization appliances do.
Free embedded SANSlide is only one of the cool new things in the IBM Storwize world. The folks in Hursley have been very busy. Check out Barry Whyte’s blog entry and the IBM Storwize product page if you haven’t done so already.
Filed under: SAN Volume Controller, Storwize V7000 |
We're all only here temporarily... 
[…] Read on here […]
LikeLike
Hi! Thank you for the post.Do you know if its going to be possible/supported to implement a Stretched Cluster Solution using SVC and IP Replication(integrated SANSlide Technology) to connect the three Failure Domains?
LikeLike
Sorry! Its not possible…Stretched Cluster and Replication using Metro/Global mirror are two different things.
LikeLike
The graph showing the red FCIP low performance line is not realistic, at least for Brocade 7800/FX FCIP devices. Not even close. The Brocade 7800/FX are and will remain the high performance product of choice for Storwize. You get what you pay for with SANslide a low performance, no compression, no IPsec solution. At the speeds demonstrated Brocade offers 4:1 compression with typical FC ingress rates of 400 MB/s for a single GE WAN connection. IPsec/VPN incurs no added cost as its included in the base switch and runs at line rate with negligible added latency, no need for firewall. The Brocade Storage Optimized TCP stack is as good or better than SANslide or Riverbed.
What I see here is… There are customers that are budget restrained and WAN connection is no problem at a price they are willing to pay. They need a freebie like SANslide to make ends meet. Then there are demanding customers that can afford Brocade 7800 4/2 to better utilize their tight WAN connection that they must share with the rest of their company by using Brocade ARL (Adaptive Rate Limiting). Instead of upping the monthly cost of BW, the investment in high performance Extension is prudent.
Also, keep in mind that with SANslide the links are Active/Passive and with Brocade 7800 the links are Active/Active coming from the Storwize. If you are going to have multiple links then using the FCIP Trunking with LLL (Lossless Link Loss) on the Brocade is truly a wonderful thing as there is no data loss across Brocade SO-TCP when a circuit is lost.
Lastly, no WANop needed with Brocade 7800. It competes in its own right against the likes of Riverbed and Silver Peak with a number of optimizations that are specific to FCIP. Keep in mind that native IP from IBM Storwize competitors may be outdone by SANslide, but not Brocade 7800 and the graph above is certainly either wrong or mislabeled with FC-IP router, or they are talking about a Cisco FCIP box.
LikeLike
Mark I think it’s important that you declare that you are an employee of Brocade. I will research some of your claims as and when I get time, and any specific evidence you can provide would be useful.
Also, do you have any comment on the generic technology issue of high latency meaning that bandwidth cannot be effectively used and which the SANSlide virtualization is designed to overcome? and are you sure that your claims for the 7800 still apply on a network with 50 ms or 60 ms latency?
If I get a chance I will also run your comments past the folks at http://www.4bridgeworks.com/
Thanks, Jim
LikeLike
Jim, Brocade did recent testing last week to verify the bogus claims of Bridgeworks, which are being promoted by IBM in the Storwize market. Everything is current. Not only does the Brocade 7800 surpass full line rate at 50-60 ms, there is no droop until after 120 ms and even then the droop out to 250 ms maintains throughput well above anything that SANslide can accomplish. Not only that, Brocade 7800 can maintain higher throughput at these latencies over SANslide with 1% packet loss thrown into the mix. This all can be proven on a properly configured 7800 using Brocade’s exclusive FCIP Trunking over a single GE connection.
Keep in mind that SANslide has not developed anything new. This is TCP technology that Nishan, CNT and Brocade have been using for years, hence, the Brocade name SO-TCP (Storage Optimized TCP). Claims of “AI” (Artificial Intelligence) is just fancy talk for “Algorithm” and who doesn’t have algorithms in TCP?
What Bridgeworks has done is develop something that other storage array companies that have native IP ports have not. I believe that the TCP stacks for native IP storage replication by other vendors, other than IBM Storwize, are quite poor in performance and robustness. That is not their core competency and TCP is a difficult technology to master. With this in mind, Brocade is a master at TCP technology with engineering having devoted substantial time and resources to it over the last 15 years.
As for mitigating the effects of latency on data throughput across a WAN connection, there are a number of techniques that can be employed many of which Brocade uses in SO-TCP. If you can find the SNA presentation by Joe White of Juniper (formerly Nishan) named TCP/IP Optimizations for High Performance WANs you will see what Brocade technology is based on.
LikeLike
OK, I’m happy that Brocade may be the king of FCIP (although personally I really liked iFCP) and that the graph might not be representative of a 7800. I have edited the blog to remove the “right away” on the end of “you don’t need to rush out and switch over to IP replication with SANSlide right away” and added some more qualifiers, especially around the 7800.
I think we can also agree that most vendors and customers are now avoiding FCIP in favour of native IP connections and that SANslide brings some very significant value to that space.
Thanks, Jim
LikeLike
I agree, iFCP was an easy protocol for many customers. The problem that iFCP had was that it could not support FICON. does not support FCR with FICON. iFCP had FCR built into the protocol. FCIP+FCR=iFCP more or less. The decision was made to separate FCR from iFCP leaving FCIP so that Brocade could effectively satisfy IBM’s mainframe customers, which we have in the largest FCIP installations in the world a few of which are using 40 Gbps connections across large distances. FCR can be enabled and disabled at will, which is a good thing. Plus, for those who don’t need it, they don’t have to pay for it resulting in lower cost FCIP.
Jim, the good news is that Brocade’s, not Cisco’s, implementation of FCIP still employs many of the tenants of iFCP in terms of how the FC frames are packaged into IP datagrams. As I mentioned before, this was superior technology developed by CNT (TankIO) and Nishan (iFCP). While maintaining the essence of those protocols Brocade was able to change the headers such that it conformed to FCIP but without the inefficiencies that are inherent in packaging each discrete FC frame independently. iFCP is not entirely dead, it was just reborn. The Brocade 7500 did not use this advanced technology, it operated the same as the Cisco boxes do today. With great inefficiency.
LikeLike
Hey Storage Buddha – have you seen what IDC has to say about SANSlide including “it is likely to give IBM a significant edge in bids for multisite solutions, something that IDC will follow with interest”? #crushlatency
http://www.4bridgeworks.com/about-us/3rd-party-opinions/
LikeLike
[…] The big guys in the Fibre Channel world can’t quite believe the speeds we run data at either. I saw some objections come up from one of these vendors who had a plethora of questions as to how we did it – our secret I’m afraid. IBM used us to fix their IP replication issue in SVC and Storwize and they got huge acceleration into the bargain, validation of our capabilities. https://storagebuddhist.wordpress.com/2013/10/08/storwize-wins-the-replication-contest-by-a-sanslid… […]
LikeLike
[…] The big guys in the Fibre Channel world can’t quite believe the speeds we run data at either. I saw some objections come up from one of these vendors who had a plethora of questions as to how we did it – our secret I’m afraid. IBM used us to fix their IP replication issue in SVC and Storwize and they got huge acceleration into the bargain, validation of our capabilities. https://storagebuddhist.wordpress.com/2013/10/08/storwize-wins-the-replication-contest-by-a-sanslid… […]
LikeLike