Purpose of this post is simple and obvious... bring back development to thick client. THANKS!
Too late! My company is already having the discussion to move to Hyper-V. I'm the lone holdout. The other techs have had it with this nonsense. That and the re-certification mess. Way to ruin things guys!
That's not the attitude that will bring on change.. Submission maybe but not change..
I'm running ESXi6 also, the performance is better BUT.
Performance is only 1 issue of many..
Security.. We are putting the management of our most critical infrastructure into the most insecure application layer we have...
resource utilization. (I really love this bloat ware... Why was it I initially used VMWare.. "OH it is small, light and fast!"
Please all elaborate.
If you build a new Version 6 vCenter Server and it takes over management of your host etc..
I'm assuming it's compatible. (I have not seen a comparability matrix yet)
The version interoperability matrix is available here and contains info with v6 releases:
As is always the case, a newer vCenter version can manage earlier ESXi releases (how else would an upgrade via VUM for example be even possible?):
What's more, you don't even need to upgrade vCenter itself to v6, since the v6 Web Client is compatible with 5.5 vCenter versions as well. You need a v6 PSC though (replaces SSO among other things):
Guys, the vSphere 6 web client is much much better. I have been beta testing v6. I really think everyone is doing themselves a dis-service by refusing to upgrade from older versions of ESX, because you are losing features and stability. And lets be realistic, the thick client is gone.
There will always be a certain percentage of users who feel obliged to adopt/upgrade production systems to the latest version on day one of its public release. These are the very same people who upgraded to 5.1 on the first day of its disastrous release, only to later realize that 5.1 contained plaintext passwords and never actually worked properly.
vSphere 6 was much more carefully prepared (we participated in the beta also), nonetheless, we will not adopt this version in production for at least 6 months from its initial public release!
The web-client does work better. But the fact remains, there are still TWO clients required for complete administrative control. This is unnecessary, and emblematic of the fact that VMware is no longer listening to its clients, it is telling them how it is!
I am certain the two client necessity will be rectified. We run two clients because we have SRM and of course VUM. Personally it does not bother me that much as I know VMware is working towards one client, and I am not in SRM and VUM all the time anyway. If I were to suggest we abandon a reliable infrastructure for Hyper-V or some other hypervisor all because I have to use 2 clients for awhile, that would seem a bit arrogant and short-sided on my part.
I certainly am not suggesting moving to another hypervisor at this point. Nonetheless, the less VMware addresses my needs, the more I will consider other platforms. It seems as if a vast amount of the R&D for the new products addresses issues that have no relevance to me. Who cares if you can apply 128 vCPU to a VM? That is a ridiculous number!
I will support the web-client when (and only when):
Until that time, all VMs will remain at VM version 8 unless there is a justifiable reason for moving to a later version. I also do not use SRM anymore, as that product represents the highest cost with the greatest complexity, for the lowest yield. Veeam has been a tremendous improvement for us!
Anyone considering going to Hyper-v to get away from 2 clients is in for a rude awakening.
Hyper-v requires 3 clients.
1. Hyper-v manager (HVM)
2. Failover cluster manager (FoCM)
3. System Center virtual machine manager (SCVMM)
SCVMM is the buggiest thing Microsoft has ever released.
I manage a Hyper-v deployment with about 100 hosts with about 30 different clusters. Nobody on our team uses SCVMM because it's so unreliable and not intuitive. We spend about 90% of our time in FOCM, 5% in HVM, and 5% in SCVMM thinking "THIS is where I SHOULD be able to manage everything, why doesn't it work?!"
I can go on for hours with my disgust for Hyper-v and SCVMM. Not a single engineer in our group of 40 or so will ever say Hyper-v is better than VMware. The guys that use to love hyper-v quickly changed their tune once they got their hands on VMware and saw how much better management is with it.
A terrible VMware product is still far superior to Hyper-v.
If anyone decides to evaluate Hyper-v, here are some things to test/consider:
1. expanding VM disk size
2. upgrading the hypervisor version, particularly with clusters
3. troubleshooting VM performance
4. frequency of patching
5. virtual networking and how to set it up with multiple VLANs
... all this discussion of Hyper-v is getting me into a bad mood.
Agreed. When we looked at, we were amazed at the shortcomings of their network configurations for VMs. I will say we Hyper-V to replicate our Vcenter. About the only things its good for. Now I am getting in a bad mood thinking about how VMware dropped heartbeat. They keep piling more and more importance on Vcenter yet kind of screw us when it comes to HA about it. We replicate Vcenter to another datacenter via Hyper-V.
Sadly this is true: Hyper-V is very nice for small business, but it cannot be compared to VMWare.
I used HyperV for about 25 hosts and did not see the need to change to VMWare.
Now we use VMWare with 3 hosts and 70 VMs and I would never go back. I do not want to praise every way it's better: is is just the better product and that why you have to pay for it.
Still the web client of 5.5 is a mess and using Flash is a real big fail. I still look forward to 6, but we will wait at least 6 months to upgrade. This is one of the two things HyperV really does better: it runs stable from the very beginning. The second thing is, that it is much easier to learn and administrate.
I am not scared hyperv can challange vmware to fast but for real vmware goes straight microsofts shit way of drop and recreate products any single release.
See vsphere 6 comes with new webclient. if we look back there was all time a web client and it got reworked all time. now we have another waste version 5.x and we all know they have to rework it at all because stay with flash and java .... seriously.
I just see 10 years of developement "waste" in the web client and also in vcenter.always just rework and recreate doing mistakes all time again and again.
VMware should be aware of the time they need to react on changes in api and documentation with their products and that 3rd party tools cant grow to stable if they change everything every time.
Why they waste time in rework a client instead of repair bugs in vaai or storage reclaim or vsan.
My keypoint is that microsoft sucks because they never get one piece of software stable, they just shit features not stable for enterprise and not suitable for small enterprise.
Same here on vcenter third architecture change in two releases. fist implement and force everyone to sso with or without multiple auth backends and totally shit implementation from rsa.
Than rework but still forced use it wothout multiple auth backends but the chance to seperate every features.
Now took away chance to split every feature because it is to complicated? What they design a high end enterprise scale solution no one in small enterprise use the features and after this they took away the chance to scale out every component because its to complicated?
I am so confused they make design for big enterprise and after this they say deployment simple for small enterprise?
In my opinion its just necessary to improve more modular components plain described and go away from "we use inventory service for cache" because thats the point why in 5.5 you cant administer storage in web client they just fake performance with short list views and cache and thats shit i hope 6.0 has a relieable view on the infrastructure in huge environments.
i will test it asap but maybe enterprise hypervisor focus on enterprise envirnments and dont try to create solutions for enterprise administered by the room maid.
We, in the VMware product development team, are looking at improving high availability for vCenter Server with highest priority. I want to let everyone know that for vCenter on Windows, we have tested Microsoft Cluster Services (MSCS) as an active-passive failover solution for protecting vCenter 5.5U2 and 6.0 environments. The vCenter 5.5 High Availability guide is available here. vCenter 6.0 guide will be available soon. If your vCenter deployment is virtualized, there's always that option of using vSphere HA along with the Watchdog feature. Watchdog monitors vCenter services for failure and attempts to restart them.
We, in the VMware product development team, are looking at improving high availability for vCenter Server with highest priority.
Literally the only availability problem I've ever seen with vCenter was either Flash crashing on desktops when trying to use it, or new desktops and servers without Flash installed at all.
This entire statement is marketing fluff - no one could care less about MSCS with this situation in play.
I will just reiterate what I posted months ago. If VMware had started developing a C# GUI with Unity3D, they would be done by now. Unity 5 is out too with a native built in GUI. Check it out VMware, it deploys to 21 platforms.... You could use the vsphere client on ios, android, linux, mac, windows, etc... or they could go the route of webgl, ALL, options are on the table AND you only have to write it once.
Wow, two whole months since the last critical Flash vulnerability; guess it's getting better. Great choice vmware.
Red Hat Security AdvisorySynopsis: Critical: flash-plugin security updateAdvisory ID: RHSA-2015:0697-01Product: Red Hat Enterprise Linux SupplementaryAdvisory URL: https://rhn.redhat.com/errata/RHSA-2015-0697.htmlIssue date: 2015-03-17CVE Names: CVE-2015-0332 CVE-2015-0333 CVE-2015-0334 CVE-2015-0335 CVE-2015-0336 CVE-2015-0337 CVE-2015-0338 CVE-2015-0339 CVE-2015-0340 CVE-2015-0341 CVE-2015-0342
Here it is: The word has come down from management that we will NOT qualify systems which depend on Adobe Flash.
It does not mater if the web-client is "improved", or "almost-as-good". We can't use it!
That doesn't mean we won't upgrade to vSphere 6, but our VMs WILL remain at VM version 8 no matter what. And we will use the C# Client to administer all of our systems.
I truly look forward to the single-client vSphere environment that doesn't depend on at least one Windows system (to run the client), but that relies on VMware creating an HTML5 client that doesn't use flash!