Re: [Users] Ovirt Memory Page Sharing or to less memory
2012/10/17 Alan Johnson <alan@datdec.com>
On Wed, Oct 17, 2012 at 5:02 PM, Mathok <mathok1@googlemail.com> wrote:
Same here with ovirt 3.1.0-2 i can't see a change in the UI. But i'm able to start more VMs and hava a higher usag of my swap partiotion so i think the change had the desired effect
Oh, does this take advantage of host swap space as well? I thought I had read oVirt could do that somewhere, but I have been reading about so many different virtualization systems over the last few months that when the RHEV 3.1 Admin Guide<https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Virtualization/3.1-Beta/html-single/Administration_Guide/index.html>did not mention swap space in section "4.2.2.2. Memory Optimization Settings Explained", I was not sure if it applied here. This is all it says at the moment:
"Memory page sharing allows virtual machines to use up to 200% of their allocated memory by utilizing unused memory in other virtual machines. This process is based on the assumption that the virtual machines in your Red Hat Enterprise Virtualization environment will not all be running at full capacity at the same time, allowing unused memory to be temporarily allocated to a particular virtual machine."
Do you have a reference handy that confirms and explains this? I put some speedy SSDs in my hosts to hopefully take advantage of this (and maybe run gluster), so I am hoping to see that in print somewhere. I'll dig it up if you don't have it handy, but a confirmation that this is your experience and understanding would make me feel better in the mean time.
let me explain my (now fixed) problem i have a ovirt host with 8 GB Ram and 12 VMs every with 256 MB Physical Memory Guarantee. When i start my two mysql-cluster vms they will use the 256 PMG plus 3,75 GB MS they have defined a max of 4 GB MS. At this poitn now the problem begins the mysql taken the hole memory in both vms and now there is no more free ram at the host and it is not possible to start more vms. (this is also speculation im not a expert) So i need for my understanding a higher rate of memory overcommit to be able to start all vms, and for that the host must use the swap to have enough memory for overcomit to start all vms and yes the higher the rate, the worse the performance. I'm not sure if this is the right or good way tho fix this problem (sorry for my bad english I hope this is understandable)
2012/10/18 Mathok <mathok1@googlemail.com>
2012/10/17 Alan Johnson <alan@datdec.com>
On Wed, Oct 17, 2012 at 5:02 PM, Mathok <mathok1@googlemail.com> wrote:
Same here with ovirt 3.1.0-2 i can't see a change in the UI. But i'm able to start more VMs and hava a higher usag of my swap partiotion so i think the change had the desired effect
Oh, does this take advantage of host swap space as well? I thought I had read oVirt could do that somewhere, but I have been reading about so many different virtualization systems over the last few months that when the RHEV 3.1 Admin Guide<https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Virtualization/3.1-Beta/html-single/Administration_Guide/index.html>did not mention swap space in section "4.2.2.2. Memory Optimization Settings Explained", I was not sure if it applied here. This is all it says at the moment:
"Memory page sharing allows virtual machines to use up to 200% of their allocated memory by utilizing unused memory in other virtual machines. This process is based on the assumption that the virtual machines in your Red Hat Enterprise Virtualization environment will not all be running at full capacity at the same time, allowing unused memory to be temporarily allocated to a particular virtual machine."
Do you have a reference handy that confirms and explains this? I put some speedy SSDs in my hosts to hopefully take advantage of this (and maybe run gluster), so I am hoping to see that in print somewhere. I'll dig it up if you don't have it handy, but a confirmation that this is your experience and understanding would make me feel better in the mean time.
let me explain my (now fixed) problem i have a ovirt host with 8 GB Ram and 12 VMs every with 256 MB Physical Memory Guarantee. When i start my two mysql-cluster vms they will use the 256 PMG plus 3,75 GB MS they have defined a max of 4 GB MS. At this poitn now the problem begins the mysql taken the hole memory in both vms and now there is no more free ram at the host and it is not possible to start more vms.
(this is also speculation im not a expert) So i need for my understanding a higher rate of memory overcommit to be able to start all vms, and for that the host must use the swap to have enough memory for overcomit to start all vms and yes the higher the rate, the worse the performance. I'm not sure if this is the right or good way tho fix this problem
(sorry for my bad english I hope this is understandable)
sorry i have no reference handy that confirm or explain this for ovirt. i
have only combined my linux and vmware knowledge and i think that it will work in ovirt a almost equal way like vmware http://www.yellow-bricks.com/2011/08/18/swap-to-host-cache-aka-swap-to-ssd/w... you can speed up the swap in a overcomit situation wit a ssd. On linux we can simply put the swap on a ssd and we have a very fast swap.
This is a multi-part message in MIME format. --------------060008020305080709020808 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Mathok :
2012/10/17 Alan Johnson <alan@datdec.com <mailto:alan@datdec.com>>
On Wed, Oct 17, 2012 at 5:02 PM, Mathok <mathok1@googlemail.com <mailto:mathok1@googlemail.com>> wrote:
Same here with ovirt 3.1.0-2 i can't see a change in the UI. But i'm able to start more VMs and hava a higher usag of my swap partiotion so i think the change had the desired effect
Oh, does this take advantage of host swap space as well? I thought I had read oVirt could do that somewhere, but I have been reading about so many different virtualization systems over the last few months that when the RHEV 3.1 Admin Guide <https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Virtualization/3.1-Beta/html-single/Administration_Guide/index.html> did not mention swap space in section "4.2.2.2. Memory Optimization Settings Explained", I was not sure if it applied here. This is all it says at the moment:
"Memory page sharing allows virtual machines to use up to 200% of their allocated memory by utilizing unused memory in other virtual machines. This process is based on the assumption that the virtual machines in your Red Hat Enterprise Virtualization environment will not all be running at full capacity at the same time, allowing unused memory to be temporarily allocated to a particular virtual machine."
Do you have a reference handy that confirms and explains this? I put some speedy SSDs in my hosts to hopefully take advantage of this (and maybe run gluster), so I am hoping to see that in print somewhere. I'll dig it up if you don't have it handy, but a confirmation that this is your experience and understanding would make me feel better in the mean time.
let me explain my (now fixed) problem i have a ovirt host with 8 GB Ram and 12 VMs every with 256 MB Physical Memory Guarantee. When i start my two mysql-cluster vms they will use the 256 PMG plus 3,75 GB MS they have defined a max of 4 GB MS. At this poitn now the problem begins the mysql taken the hole memory in both vms and now there is no more free ram at the host and it is not possible to start more vms.
In my understanding, you need to hard cap the physical memory used by the cluster VM. That will make sure all of the physical memory used by the VM will not exceed the cap, say 2G. In total, the two cluster VMs will use 4G memory at most, then the host will have 4G memory at least to start other VMs. Is that what you want?
(this is also speculation im not a expert) So i need for my understanding a higher rate of memory overcommit to be able to start all vms, and for that the host must use the swap to have enough memory for overcomit to start all vms and yes the higher the rate, the worse the performance. I'm not sure if this is the right or good way tho fix this problem
(sorry for my bad english I hope this is understandable)
_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
-- --- ?? Shu Ming Open Virtualization Engineerning; CSTL, IBM Corp. Tel: 86-10-82451626 Tieline: 9051626 E-mail: shuming@cn.ibm.com or shuming@linux.vnet.ibm.com Address: 3/F Ring Building, ZhongGuanCun Software Park, Haidian District, Beijing 100193, PRC --------------060008020305080709020808 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit <html> <head> <meta content="text/html; charset=ISO-8859-1" http-equiv="Content-Type"> </head> <body bgcolor="#FFFFFF" text="#000000"> <div class="moz-cite-prefix">Mathok :<br> </div> <blockquote cite="mid:CA+vKXs-1R_Ux8zao7CMO8N_wbejrCWTpr03xzZS6uUcTB1=fWg@mail.gmail.com" type="cite"><br> <br> <div class="gmail_quote">2012/10/17 Alan Johnson <span dir="ltr"><<a moz-do-not-send="true" href="mailto:alan@datdec.com" target="_blank">alan@datdec.com</a>></span><br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div class="im">On Wed, Oct 17, 2012 at 5:02 PM, Mathok <span dir="ltr"><<a moz-do-not-send="true" href="mailto:mathok1@googlemail.com" target="_blank">mathok1@googlemail.com</a>></span> wrote:<br> </div> <div class="gmail_quote"> <div class="im"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <div> <div>Same here with ovirt 3.1.0-2 i can't see a change in the UI. But i'm able to start more VMs and hava a higher usag of my swap partiotion so i think the change had the desired effect</div> </div> </blockquote> <div><br> </div> </div> <div>Oh, does this take advantage of host swap space as well? I thought I had read oVirt could do that somewhere, but I have been reading about so many different virtualization systems over the last few months that when the <a moz-do-not-send="true" href="https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Virtualiza..." target="_blank">RHEV 3.1 Admin Guide</a> did not mention swap space in section "4.2.2.2. Memory Optimization Settings Explained", I was not sure if it applied here. This is all it says at the moment:</div> </div> <blockquote style="margin:0 0 0 40px;border:none;padding:0px"> <div class="gmail_quote"> <div>"Memory page sharing allows virtual machines to use up to 200% of their allocated memory by utilizing unused memory in other virtual machines. This process is based on the assumption that the virtual machines in your Red Hat Enterprise Virtualization environment will not all be running at full capacity at the same time, allowing unused memory to be temporarily allocated to a particular virtual machine."</div> </div> </blockquote> <div class="gmail_quote"> <div><br> </div> <div>Do you have a reference handy that confirms and explains this? I put some speedy SSDs in my hosts to hopefully take advantage of this (and maybe run gluster), so I am hoping to see that in print somewhere. I'll dig it up if you don't have it handy, but a confirmation that this is your experience and understanding would make me feel better in the mean time.</div> </div> </blockquote> <div><br> let me explain my (now fixed) problem i have a ovirt host with 8 GB Ram and 12 VMs every with 256 MB Physical Memory Guarantee. When i start my two mysql-cluster vms they will use the 256 PMG plus 3,75 GB MS they have defined a max of 4 GB MS. At this poitn now the problem begins the mysql taken the hole memory in both vms and now there is no more free ram at the host and it is not possible to start more vms. <br> </div> </div> </blockquote> <br> In my understanding, you need to hard cap the physical memory used by the cluster VM. That will make sure all of the physical memory used by the VM will not exceed the cap, say 2G. In total, the two cluster VMs will use 4G memory at most, then the host will have 4G memory at least to start other VMs. Is that what you want?<br> <br> <br> <blockquote cite="mid:CA+vKXs-1R_Ux8zao7CMO8N_wbejrCWTpr03xzZS6uUcTB1=fWg@mail.gmail.com" type="cite"> <div class="gmail_quote"> <div><br> (this is also speculation im not a expert) So i need for my understanding a higher rate of memory overcommit to be able to start all vms, and for that the host must use the swap to have enough memory for overcomit to start all vms and yes the higher the rate, the worse the performance. I'm not sure if this is the right or good way tho fix this problem<br> <br> (sorry for my bad english I hope this is understandable) <br> </div> </div> <br> <br> <fieldset class="mimeAttachmentHeader"></fieldset> <br> <pre wrap="">_______________________________________________ Users mailing list <a class="moz-txt-link-abbreviated" href="mailto:Users@ovirt.org">Users@ovirt.org</a> <a class="moz-txt-link-freetext" href="http://lists.ovirt.org/mailman/listinfo/users">http://lists.ovirt.org/mailman/listinfo/users</a> </pre> </blockquote> <br> <br> <pre class="moz-signature" cols="72">-- --- 舒明 Shu Ming Open Virtualization Engineerning; CSTL, IBM Corp. Tel: 86-10-82451626 Tieline: 9051626 E-mail: <a class="moz-txt-link-abbreviated" href="mailto:shuming@cn.ibm.com">shuming@cn.ibm.com</a> or <a class="moz-txt-link-abbreviated" href="mailto:shuming@linux.vnet.ibm.com">shuming@linux.vnet.ibm.com</a> Address: 3/F Ring Building, ZhongGuanCun Software Park, Haidian District, Beijing 100193, PRC</pre> </body> </html> --------------060008020305080709020808--
participants (2)
-
Mathok -
Shu Ming