 |
|
VMware to make GSX no charge
Good move for VMware. GSX being free will save customers up to $2800 per server if that is the level of service they want. Good move. Like drugs..if you use, you want more..then you move up to ESX for more function and pay. I like it.
Don't confuse this move with APV. APV and VMware in my opinion are not system justifing solutions. I think you choose your platform for the value/cost of the platform. Then you add virtualization to get maximum utilization. For POWER, you can only add APV..and it's good stuff for a relatively low price. If you choose an x86 platform, you will have a choice of XEN or VMware. If you need basic support, Xen is all you need. And it is free...so let's see basic support for free or basic support for $2800. BIG choice here. So...I wonder why VMware went free with GSX??
Feb 03 2006, 08:38:00 PM EST
Permalink
|
|
Xen 3.0 is out.
Yup...Xen 3.0 is now available and it looks pretty good on paper. Are you using it? Is it good enough for enterprise mission critical work? not sure..let us all know.
Also, in the Xensouce website and announcment, it says they are working on a PowerPC version. That may be but it do not mean that version will run on a System p or a pSeries server. Remember that Xen is a type II virtualization engine and runs on the "metal". As such, it is built so it will NOT run on the Power systems hypervision. If a PowerPC version is enabled, it might run on very on POWER servers or on Apple/mac systems.
So, don't assume this means you have to compare Xen and APV to see which is better for your System p5 550. No choice... you choose APV and you get a better virtualiztion engine also. (but it will cost you $$$)
Jan 14 2006, 04:21:00 PM EST
Permalink
|
|
More virtualization features entering the market..more choices
It is rumored that HP will announce improved HPUX virtualization support (sub-processor support) and that Virtual Server will be announced to support Linux and Solaris. These are good, but I bet the marketing material around these launches will fare exceed what the announcments really mean. If you look at APV and VMware, the HPUX and VS improvements are obviously still in catch up mode.
Dec 01 2005, 10:28:00 AM EST
Permalink
|
|
Xen..VMware..VMware..Xen...Xen.. VMware
Seems like the news is always full of XEN and VMware. These are good products (well, XEN is still growing..and not out yet)...but POWER has APV now (but that's not my point)
At http://www.governmententerprise.com/news/showArticle.jhtml?sssdmh=dm4.156718&articleId=172303263 it says “What’s important is the ecosystem. How quickly [will ISVs] adopt Xen?”
Someone is missing something that has been proven over and over..if the ISVs have to change code to support XEN, the XEN will be a LONG LONG time being accepted.
Hopefully the person who said that is misinformed...or quoted out of context. With VMware and APV, you don;t ahve to change your code...XEN should be the same to be exploited
Nov 05 2005, 02:10:00 PM EST
Permalink
|
|
New VIOS feature..Ethernet takeover on failure
I think this may have been lost in the news somewhere...but the new VIOS 1.2 has feature that allows automatic failover of ethernet adapters. This is cool..just set it up, define the primary and failover ethernet adapter, and run. If the hardware adapter fails, the VIOS just switches to the alternative adapter.
No need to dual VIOSs..cool for IVM (which you can only have one of)..
...and, did you catch you can only have 4 VLANs with IVM? I don;t think that is a problem, unless you need more path connections. But since VLAN uses DMA, four should be enough on smaller servers where IVM is targeted
Nov 05 2005, 01:54:00 PM EST
Permalink
|
|
AMD Pacifica to help virtualization...duh..we got POWER...why wait?
AMD is continuing to talk about it's hardware support for virtualization..http://www.eweek.com/article2/0,1895,1877794,00.asp. Interesting isn't it. To me that means catchup..first dual core (we had that with POWER) and now hardware assist so the Virtualization support can use hardware to perform functions more efficiently. Well (1) that means VMware has to change a litte and you need to upgrade. (2) POWER system/servers have been doing this for a few years now. I wonder how far POWER and APV will advance by the time VMware/Pacifica/Vanderpool start?
just a question..
Nov 05 2005, 01:42:00 PM EST
Permalink
|
|
XEN into Kernel by RedHat
Lots of news on Redhat wanting to put help with technical support and coding to put Xen into the 2.6 kernel. That's good, but when you read more sites, your see that IBM is already working with the OS community to also assist. Seems like IBM does not make the news as much as Redhat relative to Linux...and still no XEN for POWER certification yet.
Nov 05 2005, 01:26:00 PM EST
Permalink
|
|
New VMware ESX release
VMware, www.vmware.com, is releaseing the new ESX 3. It will support up to 2 processors per partition. This is good, but I wonder why have to limit it to 2? why not let the number of processors be much larger?
Oct 21 2005, 04:35:00 PM EDT
Permalink
|
|
We announced IVM!!!
Yes, on Oct 4 IBM announced the ability to manage partitions on POWER5 servers without the need for an HMC. At first I think people are going to try to view this as a cost saver and do try and not purchase HMCs. A word of caution here: IVM in my opinion, was not created to eliminate the need for an HMC..rather it was created to support environments what were very limited in their RAS, hardware management, and partitioning requirements. In these environments, say on a p5-510, the cost of an HMC could equal the cost of the server...kind of strange if all you want to do is run two partitions. My view is that the IVM is great stuff!!...cool interface!!! remote manangement!!! But not a replacement for an HMC in all cases. If you need more than 1 VIOS or more than 4 VLANs per partition, etc..then IVM is not for you. If you are running a single system and want to only run two partitions, then IVM could be just the ticket...enable virtualization at an even lower cost!!
So, don't look at IVM as an HMC eliminator..look at it as a way to manage single systems with limited requirements.
A fly swapper kills flys very well. A gun kills elephants easily. Using a gun to kill flys is a little much; and I doubt a fly swatter will do much of anything with a herd of charging elephants...or .... use the right tool for the right job...and now we have more tools to use.
Oct 11 2005, 02:32:00 PM EDT
Permalink
|
|
Can Virtualization justify a server?
Can Virtualization justify a server?
The question may come up, can virtualization justify a server. My opinion is no. Why would you buy a server to run virtualization?? Virtualization does NO work..it manages work. However, virtualization capabilities can help decide which server and software to purchase if you have a need for or can benefit from virtualization.
Let’s say you have a few workloads running on Linux that could run on an Intel/x86 server, a POWER server, or a Sun server. If the applications and the OS’s are consistent, then virtualization could be the deciding factor. Of course there are other minor considerations like cost, performance, management, reliability, growth potential…but again, if all that is moot or equal, the Virtualization support could be the swing vote. Microsoft VS does not support Linux (yet) so that is out. VMware does support Linux but costs a lot and does not have all the features and flexibility of APV. So, if you need dynamic balancing, choice of V I/O capabilities, dynamic processors, etc…you should strongly look at POWER as the server.
At least that is my opinion.
Sep 28 2005, 12:58:00 AM EDT
Permalink
|
|
survey
Interesting..latest survey (by a famous group) says you all do virtualization to accomplish the following:
1. Reduce costs
2. Simplify the infrastructure and admin
3. Increase scalability
Aren't the last two just how you calculate the first one? Sometimes surveys are very odd.
But in this case, I agree with the results (not that anyone asked)..what do you all think?
Sep 19 2005, 01:03:00 PM EDT
Permalink
|
|
common terminology
This is starting to drive me nutz and it should you also...
The terms that companies are using to describe their function is and can be confusing. It seems different people are using the same terms to describe different functions. For example, DPLAR to IBM and VMware are different things. To say IBM can change memory, processors, and disks without a reboot is fine if you understand the DLPAR is the dedicated partition (and that is NOT what the D stands for...that is for Dynamic..even though in the share pool the resources are dynamic..um...this is even confusing in itself isn;t it?). But VMware uses DLPAR to discuss their allocation of usage in processors to groups. THis kind of equates to the shared pool (micro). But VMware cannot change allocations to a partition with out a reboot if the partition is a fixed partition...I wonder what XEN and Virtuosso call their stuff. Very confusing to compare choices. Do we need a standard here???
Sep 17 2005, 10:55:00 AM EDT
Permalink
|
|
Attendance very high at Tech U
Just back from the Linux AIX technical University and it seems to have been a great success. OVer 700 people in attendance and the response to virtualization was VERY VERY high. Labs and lectures were full...and not with tire kickers, but with real users wanting to know more. THis seems to be a good trend.
Anyone attend and want to comment on what they thought?
Sep 17 2005, 10:49:00 AM EDT
Permalink
|
|
High level sizing of APV servers
Thanks to others who have contributed this for inclusion here. This is needed information for people implementing APV since there has to be an understanding that APV does have a performance impact that is VARIABLE. But as I have said many times, APV is not about performance, it is about utlization. Having a server that is only 20% utilized says there are 80% resources available. APV is about how to use those..if you loose 5% you still have 75% available to use...
The secret to sizing a server with Advanced Power Virtualization (APV) is to size CPU.s for the average workload. This is in contrast to a traditional server (or LPAR), which is sized for peak workloads.
Advanced Power Virtualization allows a physical server to run multiple partitions, sharing CPUs, boot disks, network and SAN adapters. Sharing uses the hardware more efficiently. Consequently less hardware is needed for the same workload.
The attached example illustrates how I would size an APV server to support the equivalent workload of 10 p510.s. I.ve quantified some of the benefits, which in this example reduced hardware costs in excess of 20%, and reduced power, cooling and rack space by over 80%. All considered, APV more than pays for itself.
Standard disclaimer: each situation is different, your results may vary.
Example
Assume there are 10 p510.s, each having 1x1.65GHz CPU and 1 GB memory. Each p510 is sized for the peak loads, with an expected average CPU utilization of 20% (Gartner.s industry average utilization rate for Unix servers). Each p510 has dual Ethernet and dual boot disks.
APV Sizing Methodology
CPU Sizing and Model Selection: The CPU.s for APV are sized for the average combined workload of all the partitions, plus a shared free pool of CPU resources for peak processing. A small VIO partition is also required to support shared boot disks and Ethernet adapters.
The sizing metric is IBM's rPerf performance rating. The p510 has a rPerf rating of 5.24. The expected average utilization is 20%. So the aggregate average rPerf for 10 servers, and VIO server would be:
p510 rPerf: 10.48 (10 servers)*(5.24 rperf/server)*(20% util)
VIO rPerf: 1.0
------------------------------------------
Average rPerf = 11.48
Additional capacity is required for the shared CPU pool for peak processing. As a first pass, I size the free pool by specifying the overall server utilization. A reasonable number would be 60%. This results in a overall rPerf rating of 19 (11.48/60%). A good APV server would be the p550 which has an rPerf rating of 19.66 (4x1.65GHz).
Memory: The total memory requirement is 10 GB (10 servers * 1 GB/server). Rounding up to 12 GB provides the additional memory for partitioning overhead (700MB), and matches available memory hardware increments.
Boot disks: The client partitions will boot off shared .virtual disks. located on the VIO server. A total of 8-72 GB drives will be sufficient. Two disks (mirrored) will be allocated to the VIO server operating system, and the remaining 6 disks (420GB) will support the AIX requirements of the 10 partitions with 100% spare capacity.
This assumes 20 GB disk per AIX partition and modest I/O rates. AIX will comfortably fit on 10 GB disk (OS + paging), or 20GB with redundancy. .Modest I/O. rates means no paging or heavy I/O as in a database.
Network: Client partitions will share 2 redundant GigaBit Ethernet adapters located on the VIO server.
Standalone vs APV Configuration Comparison
P510 P550
Physical Servers 10 1
CPU's (per server) 1 x 1.65 GHz 4 x 1.65 GHz
Memory 1 GB 12 GB
Boot disk 2x72 GB 8x72 GB
Ethernet 2 2
HMC 1 1
rPerf 5.24 19.66
Expected CPU Util. 20% 58%
Results
P510 Alternative P550 APV Alternative APV Savings
List Price $72,370 $55,528 25%
CPU's 10 4 60%
Max power (watts) 6000 1100 82%
Max cooling (btu) 20460 2557 87%
Ethernet ports 20 2 90%
Rack Space (height) 35 7 80%
In this example, APV reduced hardware costs by $17,000 or 25%. Other tangible savings include power, cooling and rack space. Depending on charge back rates, this translates to savings of around $10,000 per year.
Another significant benefit is the software savings, which is often overlooked in the hardware purchase. Many software applications (Oracle, DB2, Websphere) are priced based on the number of CPU.s. In this example, APV would reduce this software cost by 60% because we.re using 4 versus 10 CPU.s. To provide an order of magnitude savings, assume the software costs $20,000 per CPU. If so, the savings would be $120,000 - more than double the cost of the hardware itself! In essence, the hardware is paid with the software savings.
Finally, APV flexibility reduces the sizing risk compared to standalone servers. CPU resources are .load leveled. dynamically. Undersized partitions will dynamically grab CPU cycles from the shared free pool while oversized partitions will donate unused cycles back to the free pool.
Memory resources can be redistributed between partitions. Memory granularity is in MB, and the reallocation is dynamic.
Conclusion
When using Advanced Power Virtualization, servers are sized based on the aggregate average workload, plus a smaller shared pool of CPU.s for peak processing. APV requires a paradigm shift in thinking, but is well worth the effort.
Sep 13 2005, 12:31:00 PM EDT
Permalink
|
|
|
| S | M | T | W | T | F | S | | | 1 | 2 | 3 | 4 | 5 | 6 | | 7 | 8 | 9 | 10 | 11 | 12 | 13 | | 14 | 15 | 16 | 17 | 18 | 19 | 20 | | 21 | 22 | 23 | 24 | 25 | 26 | 27 | | 28 | 29 | 30 | | | | | | | | | | | | | | Today |
|
Virtualization
