Présentation de SCVMM 2012

Vous êtes ici : Accueil > Articles

Publié le : 11/08/2011 vers 22h
Mise à jour le : 08/10/2011 vers 0h
Catégories :
  • Hyper-V

Auteur(s) :
Loïc THOBOIS (Membre depuis le 04/09/2007)
Société : EGILIA
Fonction : Direction filière technologies Microsoft
Contactez cet auteur - Affichez les ressources de cet auteur
William CARON (Membre depuis le 10/06/2009)
Société : EGILIA
Fonction : Formateur
Contactez cet auteur - Affichez les ressources de cet auteur
Michaël ESSON (Membre depuis le 15/04/2009)
Société : EGILIA
Fonction : Formateur
Contactez cet auteur - Affichez les ressources de cet auteur

Introduction à SCVMM 2012

SCVMM 2012 est comme son nom l’indique la nouvelle version de l’outil de la suite System Center permettant de gérer les environnements virtuels au sein du SI.

Ces dernières années, la virtualisation de serveur s’est introduite aussi bien dans les petites entreprises que dans les plus grands datacenter. Les usages ont naturellement évolués et les besoins de gestion et de contrôle de l’infrastructure fortement augmentés nécessitants des outils d’administration et de supervision internes aussi puissants que ce que l’on trouvait uniquement chez les hébergeurs il y a encore quelques temps.

Ainsi on observe de plus en plus la mise en place d’un service dédié à l’administration de l’infrastructure virtuelle qui aura pour tâche de livrer les ressources (puissance de calculs, stockage, bande passante réseau, fonctionnalités informatique, …) aux autres services lorsqu’ils en ont besoin.

On assiste donc à la naissance du cloud privée (interne) qui va réduire de manière importante les délais de livraisons des serveurs un peu comme ce que nous connaissions à l’époque des services de gestion du stockage. Avec ce nouveau mode de fonctionnement, le délai de mise à disposition d’un serveur passe de plusieurs semaines (dimensionnement des serveurs, commandes, délai de livraison, mise en conformité du serveur et déploiements des composants logiciels validés en internes, …) à seulement quelques jours (Les serveurs sont déjà présents et supervisés, les images logiciels sont pré-validées, …).

Ce modèle va permettre de faciliter la refacturation des ressources aux différents services de l’entreprise transformant le service informatique en centre de service plutôt qu’en centre de coût.

SCVMM 2012 s’inscrit naturellement dans cet esprit en proposant des améliorations facilitants le déploiement d’une infrastructure virtuelle puissante et agile.

Les fonctionnalités présentées dans cet article sont basés sur la beta de SCVMM 2012 et nous pouvons espérer une version Release Candidate aux alentours de septembre.

Votre commentaire:

Votre pseudo :

Votre commentaire :

Saisir les caractères ci-contre dans le champ suivant :
 Rafraichir

PseudoCommentaire
jVrcpBIHnFD
15/05/2012 21:09:35
lloBjh usgnsqepmaeg
OAsFfBiJrkj
14/05/2012 22:00:13
i8wyMG aoxgknisagpu
SuelOWUAMPouHXRsR
14/05/2012 20:36:52
vNmd0G azxaxwwdprbp
pDBFgATAWfkkfoZT
14/05/2012 02:44:12
Hi Duncan, I have successfully used your mtohed to deploy ESXi4 on USB keys. The only way to run VMs is to attached a data store via iSCSI, NFS or local drives. Personally I want to use the USB key itself to store my VMs but could not find out a way… Any idea ?
dHAOpMGr
13/05/2012 06:26:48
9ga929 wovjjhpcvtkz
ebGnSNxunSa
13/05/2012 05:35:28
emIdbo uimtjwommrbe
DvkaPBeZXxclKqD
13/05/2012 04:53:15
CSkXEU njubunnoaxht
oPdmHehcgIj
12/05/2012 18:33:02
Good to see real expetrise on display. Your contribution is most welcome.
CVVwSDvsGjGkX
12/05/2012 17:45:38
Duncan,I appreciate eeovyrne that has come here has been from a major VMware background, or even from VMware themselves and are obviously soured by the fact that the competition might have something you don't. Nowhere do I say Microsoft *must* be used or that VMware is absolute rubbish, I'm simply making a statement that Hyper-V caters for easier managability in this instance than does VMware, fact otherwise it wouldn't have caused such a stir.When you say you work for Enterprise Customers, I don't deny that, nor do I question it. I come from a major enterprise background myself, and have implemented numerous systems all over the world for major corporates. Now, TPS, memory overcommit or ballooning, no matter how you want to look at it, opens a potential can of worms when it comes to security. By finding commonalities and streamlining them in to one, essentially what TPS is doing, allows for a virus or root kit potentially to find its way in to one of the commonalities and then propogate itself amongst all of the servers in question. Furthermore, when you overcommit memory, you're doing it on the hopes and pretenses that all of the servers you've got in the infrastructure (obviously on any given host) aren't ever going to be fully utilised at any given time, because if they are and you don't have enough memory, or you've over committed it in the wrong ways, your infrastructure comes to a grinding halt as that is the way RAM works. Interestingly enough,, the following is verbatim re memory:Avoid frequent memory reclamation. Make sure the host has more physical memory than the total amount of memory that will be used by ESX plus the sum of the working set sizes that will be used by all the virtual machines running at any one time.Which actually negates memory overcomitt altogether, because if you want more physical memory than the total amount of memory that will be used by ESX plus the sum of the working sets, why overcommit?Might want to ask yourself though how you ended up in this situation as any environment needs to be monitored / managed and clearly no one bothered to over the last months.When a client is using their infrastructure on a Friday and turns the servers off to move to a new environment on a Monday, they're obviously being monitored, else they would have called myself or another IT provider prior to the move. I'd think of rephrasing your question as it really doesn't seem to make much sense No where did I ever mention the system was offline for months or that it was unresponsive for an extended period of time. The article reads:Today I got called to a client’s site who recently moved from premises A to premises B. This meant they turned off their VMware ESX 3.0 server for the first time in a while. When they relocated they couldn’t authenticate nor could they get to their file/print server.They moved site, hence all of the physical hardware was moved. It worked when they turned it off last week before moving and this week, it didn't. That isn't monitoring/management.
ZnLmfTUnaPFxVkkgt
12/05/2012 09:29:23
It is really uasiredtimnteon of the trouble VMware get himself into now.Most customers can stay with existing ESXi 4.1 for a long time, and will calculate the cost of alternatives. So VMware will not get more money from the new licensing for sure. But it ALREADY made many people to put VMware extension plans on hold and look for alternatives, which DO EXIST NOW.In addition, 8 GB vRAM enrollment on the free version is just a joke. I saw many cases when people moved thru- free ESXi- Essential license- Standard or enterprise license.They knew that ESXi is fully functional system with limited management so they can always run it if they can't temporary get budgeting or if they want to demonstrate that VM strategy do work in their environment. Now this pass is broken; no one in good mind will use ESXi with 8 gb of vRAM per cpu (remember, 32 GB RAM is norm even today, 64 and 128 will be norm in 1 2 years).So, Vmware did not get any benefits from new model (people can just stay with 4.1 for 3 5 years; ESX5 _does not have any interesting features AT ALL! NO ANY! See the chart - almost nothing interesting for average enterprise! Few minor improvements in SAN access, and that's all. Except for a huge clouds_) but it effectively promoted customers to try alternatives! What a shame! Competitors can applaud!
APhfbRtLB
12/05/2012 07:09:42
Hi Christof,Unfortunately there is currently no SMI-S predvior for HP P4000. There is one for EVA and the next one coming up is for 3PAR (probably May timeframe).HP told me not to expect an SMI-S predvior for P4000 or P2000 within 12 months. Maybe if plenty of customers start to complain about this, they might change their mind and get one earlier. So until then, you'll have to provision storage via SAN/iQ management software. If you do that, you can still take advantage of bare metal deployment and cluster creation from VMM 2012 and all the other goodies.Cheerz, Hans
Nos partenaires : www.bestof-microsoft.com | www.exchangedump.com | www.exchangedump.com | www.formation-exchange.com | www.formation-management-dif.com | www.formation-mcpd.com | www.formation-pmp.com | www.formation-seven.com | www.formation-windows-7.com | www.formation-windows7.com | www.formation-windows-seven.com | www.certification-mcpd.com | www.certification-windows-7.com | www.consulting-exchange.com | www.e-managementinstitute.org | www.certification-mcitp.com | www.formation-mcsa.com | www.bestof-exchange.com | www.exchange-consulting.com | www.bestof-voip.com | www.formation-cisco.com www.formation-mcitp.com.com | www.laboratoire-exchange.com | www.messagerie-consultant.com | www.bestof-windows.com | www.bestof-sharepoint.com | www.messagerie-consultants.com | www.formation-mcts.com | www.consultant-messagerie.com