Archive for June, 2011

Busy week-end preparing MondoRescue and a concert


It was very sunny this week-end. But I didn’t had really time to benefit from it. I wanted to publish MondoRescue as I was in need of a fixed version for consulting I do around it this week at a customer site. So my full Saturday was fixing 5/6 various bug reports + mailing list reports that were the most critical IMO and important to have on top of the one already htere since last release in April.

I know there are some remaining points to handle, but you always have, and sometimes you need to draw the line, saying it’s better, even if not perfect and deliver it. I hope users will understand and be still happy with that version.

Then it was time to prepare tonight’s concert Sunday, and also prepare my attempt to improve my carrier path inside HP and hopefully reach the next level, which is in my case not easy at all.
But I can leave happy now: the version is here, and the concert was a success, especially the Salve Regina of Victoria 😉

Third day at the EMEA Red Hat partner Summit


Time again to attend the Keynote !

Werner Knoblich eVP Red Hat EMEA, covered as usual the latest results from Red Hat in EMEA:

  • YoY 34% increase
  • 3% increase on the indirect to reach 63%
  • 155 more new net people in EMEA reaching 1000 (only 100, 8 years ago !)

Werner mentioned the future role of virtualization, insisting on the importance of the openvirtualization alliance to avoid a future MicroSoft to be created, and the commitment of IBM, HP and Intel in it.

Rhys Austin,Senior Business Development Manager for HP Industry Standard Server (ISS) EMEA, was presenting the instant-on enterprise with HP CloudSystem as next keynote. He started insisting on the importance of the channel for HP, as it is for Red Hat.He mentioned that the Business department is today outpacing the IT department. So Cloud is thought to be helpful in that approach, including buying external resources to the company. Concerns around cloud are in order: Lock in, Performanace, Availability and Security; and external vs internal integration.
Rhys then described the multiple aspects a cloud architecture should cover, depending also on the user’s point
of view, and presented how HP ClouSystem could give answers to all these feature requests and concerns.

Rhys also presented HP Cloud Maps, aligned with Red Hat Cloud strategy to accelerate application deployment and create cloud services (RAC and JBoss already provided). And Rhys detailed how Red Hat products fit into HP CloudSystem.

He finally exposed that partners need to question their customers’ cloud strategy to not risk loosing business
and be positioned in that new play either as provider, aggregator or broker. HP cannot do that without partners.

Joel Berman, VP Global Field Marketing at Red Hat, was kind enough to mention how our Solution Center in Grenoble and the joint Open Source Solutions
that I initiated there between HP, Intel and Red Hat could help customers and partners make more business w
ith us ! But that was not the nicest mention for us that day 😉

Craig Muzilla, VP of the Middleware Business Unit at Red Hat, then went on stage to describe where Red Hat is on the JBoss side, in a session called “JBoss Everywhere – Better for the Enterprise. Ready for the Cloud”. He insisted on how JBoss is now used strategically by large companies such as SFR, NYSE, Allianz, …

Red Hat announced JBoss Enterprise Application Platform 6 (supporting JEE 6, and CDI contributed to JCP)
Red Hat announced also JBoss Enterprise Data Grid 6 (foundation for Cloud computing)

Craig went on describing the future of JBoss around mobility, cloud and SOA. He also covered MRG and BRMS.

We then had a very fruitfull meeting with Jim Totton and Joel Berman, which continued on our booth as well with Lars Hermann. After a bit of booth duty, it was time again to deliver our second session around best practices for RISC/Unix to RHEL on HP Xeon based ProLiant migration. You can get ideas of the topics we covered by reading the excellent blog article of Arnaud who is leading that activity at WW level for HP.
We had less people than the day before, but hopefully this will lead to some concrete further on contacts around concrete migration opportunities !

Again a dense day, but it was not the end, as we still had to attend the dinner at the Guiness StoreHouse, where Red Hat announces their prices for the best partners of the year, and organize a fun evening with the 700+ partners representatives that are attending.

And I must confess that this part of the day was the most interesting for us !! More on that in the next article.

Second day at the EMEA Red Hat partner Summit


So time to go back to the keynotes again !

Paul Cormier, eVP of Products and Technolopgies at Red Hat, started with a bit of IT history (’80 locked with Dec and Sun, ’90 locked with Microsoft, ’00 Open Source with Red Hat, ’10 Red Hat Cloud) then we went on explaining the Fedora community and linkage to RHEL, as well as JBoss and RHEV foundations. And Cloud foudation (with Cloud Forms and Open Shift as seen yesterday). I’d have expect a bit more of content on the Cloud aspects at Red Hat and its infrastructure.

Jean Staten Healy of IBM then did its yearly advertising campain again with their now famous 1B$ dedicated to Linux… in 2001 ! Ok, to be fair she also mentioned a lot KVM and the Open Virtualization Alliance whose HP is also a governing member. Of course take it with a grain of salt, as I’m working for a compettor, but I was really not impressed.

St Boniface was the starting part of the next talk of Ron Tolido of Cap Gemini, as an evangelist. OSL evangelism shouldn’t be needed anymore (being concretely in front of customers regularly, believe me this is just plain wrong and still has to happen !!).
Cloud has the same issue for him. His talk is on benchmark and the economics of total open source and the cloud. He was involved in Business Technology Agora sessions, which goal is to understand what business expects from IT. Business want to spend more budget on the higest part of the stack rather than on what he calls the invisible infrastructure which is for him boring (again, that’s probably I’m not a business consultant 😉 as I strongly disagree here again !).
All in all, no more IT engineers are needed and you “just” need to glue FLOSS components to have a great IIT. He was the one who bought the dog from Joel Berman joke on the IT consultant !!

I think it’s really a pity today to have such attitude to bash the guys who make the IT run really. Because at the end of the day, when it’s not working, who do you call ? Your favorite call center (where ever it happens to be ?) or your beloved system engineer (if you still happen to have one around to help you) ?

So I was quite happy to be able to have a pause with the Drum Café afterwards, which allowed us to beat on something at least 😉 Best part of the morning definitely !

We then had a very interesting meeting with Werner Knoblich, Gus Robertson and Robert Eiselt in order to sync up on our joint HP, Intel and Red Hat activities.

Some booth duty was planed for the luch period, before I attended a Lab on cgroups.
This is a technology which was really needed for Linux, when compared to other enterprise Unices. But the way it has been developed and integrated make it very powerful, and one of the most promising, probably after SELinux and IPtables. The ability to control resources used by what ever process is fine grained and gives lots of power to the administrator to avoid abuses, limit consumption, … up to the creation of containers !

I was not able to attend the whole Lab, due to another important snc meeting we had with the
Red Hat Regional managers (Phil Andrews, Franz Meyer, Brian Cornell). Very intereting as well. Then after some booth duty again, we delivered our first session, which was attended by some 30 persons and after that it was time to move on to the sponsor dinner and we had again some more interesting discussions with people from Nimbula and Red Hat which ended up very late during the night.

First day at the Red Hat EMEA Partner Summit


Arrived in Dublin the 5th of June at 6:00PM local time for the Red Hat EMEA Partner Summit 2011, we did the booth setup with my colleague and friend Arnaud Meurant in less than 1 hour with a demo DL 980 G7 unit and a working DL 380 G6 with RHEL 6.1, KVM and SR-IOV demos thanks to Intel Gigabit cards that my other colleague and friend Jean-Marc Andre had build up.

We then attended the keynote sessions of the evening.

Petra Heinrich VP Partners and Alliances EMEA for Red Hat said, not without humour, that the sun was now declining and clouds coming, as usual during the Red Hat Partner Summit !
Red Hat’s own Ronan Kirby then made a presentation of Ireland and Old Ireland explaining the mixed origins of the country (Celts, Vivkings, Normans, Brits, …).
He metioned the 960 multinationals employing 138000 people in R&D and IT.(compared to the 4 Million people in the country)

Joel Berman, VP Global Field Marketing, introduced the ecosystem of partners (16 sponsors helped make that event happened). And this year HP was Gold sponsor for the first time !
Joel always make a very relax but still very professionnal introduction to sessions (not counting some ecellent jokes on the IT world). One sentence I noted from his first speech about Cloud was: “You don’t call it fog computing because marketing guy are not that stupid”.

Jim Whitehurst, Red Hat CEO, then made the keynote, starting by showing how new communication means are supporting the movement in Egypt, Tunisia, … People are looking for transparency and accountability. Social changes are driving the political changes. Where and how innovation occurs has changed.

He underlined that there is Way too much hyper around cloud. It’s a paradigm shift as large as moving from the mainframe to client server. Cloud is important due to the fact it is a user driven innovation. At Google, Facebook, … Open Source helped users made what they wanted. No vendor was doing that. And they can’t control it. And no one knows where it goes.

One of the area supported by cloud is collaboration. Engineers of Google, Facebook share technologies. They first need to improve techno, doing what is right, before fighting with each other.
Then it’s transparency. The wining model is the best architecture of participation. Not only the best technologists. Red Hat doesn’t need to have everything at hands, but the right relationships to create technology that will solve users problems.
Finally choice. From CIO’s perspective paying up front to hope it works is painful. So they feel locked and frustrated. Each provider will have to show where they bring value.

All these points are also qualities of Open Source.
In a world of cloud the HyperVisor is becoming more important than the OS (as it was before). And the choice is not feature/functionality but whether we create the next Microsoft, or favour collaboration around open alternatives allowing a vibrant ecosystem.
Apps need to be written once and run on every cloud, not locked on one.

Jim did a very enlightening talk, which is bringing a real vision, and not only a bunch of empty statements. I think his contribution to make Red Hat a successful company was visible to everybody attending.

Scott Crenshaw, VP Cloud BU at Red Hat, then took over. For him, this year Cloud will move from PoC to production. However customer do not have more budget (+4% max) so yhey’ll move money from old SW to cloud SW. Red Hat is here to help their partners winning business and money in the change.
He then presented CLoudForms (IaaS) and OpenShift (PaaS) and the Open Virtualization Alliance whose HP is a governing member.
Scott then promoted cloud as lowering cost and transforming IT speed and agility.
Technology issues are much simpler in cloud than the political and cultural (buying online instead of internaly). Cloud also presents threats to xSP (disintermediation, pre-built solution reducing need of integrators). Red Hat commitment is to help partners becoming expoerts so they are the one in front of their customers. Because partners already know their customers, especially around verticals solutions.

I found it was difficult to come after Jim to speak again of cloud, where he put the bar very high while here we were back to more classical concepts and content. I was lucky enough to also be introduced afterwards more precisely into CloudForms and OpenShfit, so I know now what is behind, which wasn’t completely the case after this presentation.

Then after multiple interactions on the HP/Intel booth till 11:00PM, it was time to go back to hotel and finish the presentation for the day after 😉

Mageia 1 is now out


As promised, Mageia 1 is now out. Well, it’s not that often that a new large distribution sees the light so there are reasons to celebrate !

And one way for me was to upgrade/install some computers with it. So I reinstall my daughter machine (everything on an NFS server + LDAP + automount). 20′ later + a flash 64bit plugin install + some packages reviewed, she was just at work again, with the new distribution. painless 🙂

Then I upgraded one of my laptops. That wasn’t as easy. I PXE booted it, and it upgraded most of the packages. But when rebooting I realized that there were still 800+ packages to do. So I used urpmi --autoselect to finish the job. It could be linked to some additional install + PLF by the way. 2 run after (due to that I guess) I had everything handy. 200+ mdv2010 packages are still around, so I’ll have to push some of the one I really use in Mageia now !

I also started to check that whay I want to do with that machine was still working. was regenerated without issue (there is a new perl warning on require I’ll have to fix).

The I tries MondoRescue and encoutered 2 problems, that I reported, due to DocBook generation. The first one was around incomplete catalogs for DocBook 4.1 and the second one is with the jadetex package which seems incomplete/incoherent

Outside from that, it seems to work pretty well globally. That consumer laptop seems to still have issues with its USB 3 hotplug, but that was already not working correctly with Mandriva 2010.2. All in all a smooth upgrade. Now I will have to do the same for my other machines, destops and servers, both at home and at work, but I’ll make some further checks during my upcoming travels beofre laucnhing the big bang operation in 2 weeks. Probably a couple of updates will have landed in between, aloowing me to work even more easily with it.

I really recomend that you test it … and adopt it !