FCoE Might Improve Storage but Will Complicate IT and Cause Team Disruption

I was think­ing over the integ­ra­tion of an IT Engineering team to provide a cloud com­put­ing ser­vices. While dis­cuss­ing team respons­ib­il­it­ies and oper­a­tional “edges” I real­ised how divis­ive and dys­func­tional FCoE will be to a team. The impend­ing threat of FCoE hasn’t really receded, but the good news is that it isn’t going to arrive nearly as fast as pre­vi­ous. It is the end of September and the FCoE and CEE stand­ards are not show­ing any signs of finishing.

Team Integration

It is becom­ing clear that the impend­ing forced integ­ra­tion of Storage Engineers to work with Network Engineers is going to force a cul­ture clash of large pro­por­tions. Generally, the tech­no­logy gap from FC to Ethernet is going to be sub­stan­tial for many because they have come from a Server background.

That is, a sub­stan­tial por­tion of Storage teams were Server people who re-​​skilled or spe­cial­ised in Storage in the early phase of the FC mar­kets devel­op­ment. FC is a com­plex and dif­fi­cult pro­tocol to mas­ter and these people have real know­ledge and skills, but this know­ledge is not trans­fer­able to Ethernet switch­ing. They will need to learn new skills. A trans­ition can be difficult.

Consider the cur­rent team integration

The Network and the Server team have a demarc­a­tion that is based on IP. The com­mon point of “can you ping it” it usu­ally a clear demarc­a­tion¹.

The Server and Storage team have a demarc­a­tion that is based on FC. Same idea. So a Venn dia­gram looks like this:

This reflects that the Network and Storage teams don’t have much need to communicate.

Add FCoE — looks messy

The first deploy­ments of FCoE will cre­ate inde­pend­ent net­works that only carry FCoE traffic. The sales pitch of con­verged eth­er­net will get a lot of lip ser­vice, but stor­age people are quite def­in­ite about not shar­ing the data and stor­age net­works for “lack of reli­ab­il­ity” etc.

Which is to be expec­ted, but it won’t take long for IT man­age­ment to force the con­ver­gence issue to save money, both cap­ital and oper­a­tional. At this point, FCoE makes the team inter­ac­tion look like this:

So I can look for­ward to meet­ings where there are end­less debates, dif­fer­ent points of view, and vastly dif­fer­ent require­ments. Change con­trol is going to be a nightmare.

The iSCSI point of view.

If you take the view that iSCSI can provide Storage over IP (which it cer­tainly can), then team inter­ac­tions are going to look more like the following:

The Pain Point and its prob­able outcome

I can see the FCoE option is going to cre­ate sig­ni­fic­ant per­son­nel fric­tion in the future, the com­bin­a­tion of Ethernet and Fibrechannel will cause demarc­a­tion dis­putes that will take time, exper­i­ence and man­age­ment resources to work through. The bene­fits of Fibrechannel will need to be enough to out­weigh this or man­age­ment will stop invest­ing in the tech­no­logy and it will wither or stagnate.

The ques­tion is: Will that happen ?

Cisco and Brocade are bet­ting on FCoE and the rest of the industry is fol­low­ing along like pup­pies. FCoE sup­port­ers point to vendor announce­ments and their pre­lim­in­ary tri­als. No one is point­ing to IP Storage fea­tures that are get­ting less focus but are equally stra­tegic. After all, old and mature tech­no­logy does not make a good story.

Conclusion

I’m not look­ing for­ward to the clash of cul­tures with the Storage team. It going feel like a repeat of the main­frame net­work­ing, or Token Ring vs Ethernet, from years gone by. I pre­dict a lot of money wasted in team meet­ings arguing about point­less top­ics, and polit­ical man­oeuv­ring, and man­age­ment involve­ment that isn’t going to do much to improve anyone’s life.

It seems inev­it­able though.

Footnotes

1. okay, so per­form­ance, QoS, band­width aren’t con­sidered in this sim­pli­fic­a­tion, but run with the idea [back]

Please rate this post:

(No Ratings Yet)

 Loading ...