Il giorno 18 dic 2016, alle ore 18:05, rajatjpatel <rajatjpatel@gmail.com>= ha scritto: =20 Hi Alessandro, =20 Right now I have 2 physical server where I have host ovirt these are HP pr=
--Apple-Mail-F8806A8F-B8B6-48D9-BC6C-8A2672E65957 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Hi, oh, so you have only 2 physical servers? I've understood they were 3! Well, i= n this case ceph would not work very well, too few resources and redundancy.= You could try a replica 2, but it's not safe. Having a replica 3 could be f= orced, but you would end up with a server with 2 replicas, which is dangerou= s/useless. Okay, so you use nfs as storage domain, but in your setup the HA is not guar= anteed: if a physical machine goes down and it's the one where the storage d= omain resides you are lost. Why not using gluster instead of nfs for the ovi= rt disks? You can still reserve a small gluster space for the non-ceph machi= nes (for example a cinder VM) and ceph for the rest. Where do you have your c= inder running? Cheers, Alessandro oliant dl 380 each server 1*500GB SAS & 1TB *4 SAS Disk and 1*500GB SSD. So= right now I have use only one disk which 500GB of SAS for my ovirt to run o= n both server. rest are not in use. At present I am using NFS which coming f= rom mapper to ovirt as storage, go forward we like to use all these disk as = hyper-converged for ovirt. RH I could see there is KB for using gluster. Bu= t we are looking for Ceph bcoz best pref romance and scale.
=20 <Screenshot from 2016-12-18 21-03-21.png> Regards Rajat =20 Hi =20 =20 Regards, Rajat Patel =20 http://studyhat.blogspot.com FIRST THEY IGNORE YOU... THEN THEY LAUGH AT YOU... THEN THEY FIGHT YOU... THEN YOU WIN... =20 =20
On Sun, Dec 18, 2016 at 8:49 PM, Alessandro De Salvo <Alessandro.DeSalvo@= roma1.infn.it> wrote: Hi Rajat, sorry but I do not really have a clear picture of your actual setup, can y= ou please explain a bit more? In particular: =20 1) what to you mean by using 4TB for ovirt? In which machines and how do y= ou make it available to ovirt? =20 2) how do you plan to use ceph with ovirt? =20 I guess we can give more help if you clarify those points. Thanks, =20 Alessandro=20 =20
Il giorno 18 dic 2016, alle ore 17:33, rajatjpatel <rajatjpatel@gmail.co= m> ha scritto: =20 Great, thanks! Alessandro ++ Yaniv ++=20 =20 What I want to use around 4 TB of SAS disk for my Ovirt (which going to b= e RHV4.0.5 once POC get 100% successful, in fact all product will be RH ) =20 I had done so much duckduckgo for all these solution and use lot of refe= rence from ovirt.org & access.redhat.com for setting up a Ovirt engine and h= yp. =20 We dont mind having more guest running and creating ceph block storage a= nd which will be presented to ovirt as storage. Gluster is not is use right n= ow bcoz we have DB will be running on guest. =20 Regard Rajat=20 =20
On Sun, Dec 18, 2016 at 8:21 PM Alessandro De Salvo <Alessandro.DeSalvo= @roma1.infn.it> wrote: Hi, having a 3-node ceph cluster is the bare minimum you can have to make i= t working, unless you want to have just a replica-2 mode, which is not safe.=
Il giorno 18 dic 2016, alle ore 17:07, Yaniv Kaul <ykaul@redhat.com> h= a scritto: =20 =20 =20 On Sun, Dec 18, 2016 at 3:29 PM, rajatjpatel <rajatjpatel@gmail.com> w= rote: =E2=80=8BDear Team, =20 We are using Ovirt 4.0 for POC what we are doing I want to check with a= ll Guru's Ovirt. =20 We have 2 hp proliant dl 380 with 500GB SAS & 1TB *4 SAS Disk and 500G= B SSD. =20 Waht we are done we have install ovirt hyp on these h/w and we have ph= ysical server where we are running our manager for ovirt. For ovirt hyp we a= re using only one 500GB of one HDD rest we have kept for ceph, so we have 3 n= ode as guest running on ovirt and for ceph. My question you all is what I am= doing is right or wrong. =20 I think Ceph requires a lot more resources than above. It's also a bit= more challenging to configure. I would highly recommend a 3-node cluster wi=
It's not true that ceph is not easy to configure, you might use very ea= sily ceph-deploy, have puppet configuring it or even run it in containers. U= sing docker is in fact the easiest solution, it really requires 10 minutes t= o make a cluster up. I've tried it both with jewel (official containers) and= kraken (custom containers), and it works pretty well. The real problem is not creating and configuring a ceph cluster, but us= ing it from ovirt, as it requires cinder, i.e. a minimal setup of openstack.= We have it and it's working pretty well, but it requires some work. For you= r reference we have cinder running on an ovirt VM using gluster. Cheers, =20 Alessandro=20 =20 th Gluster.
Y. =20 =20 Regards Rajat=E2=80=8B =20 =20 _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users =20 =20 _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users =20 --=20 Sent from my Cell Phone - excuse the typos & auto incorrect =20 =20
--Apple-Mail-F8806A8F-B8B6-48D9-BC6C-8A2672E65957 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable <html><head><meta http-equiv=3D"content-type" content=3D"text/html; charset=3D= utf-8"></head><body dir=3D"auto"><div></div><div>Hi,</div><div>oh, so you ha= ve only 2 physical servers? I've understood they were 3! Well, in this case c= eph would not work very well, too few resources and redundancy. You could tr= y a replica 2, but it's not safe. Having a replica 3 could be forced, but yo= u would end up with a server with 2 replicas, which is dangerous/useless.</d= iv><div>Okay, so you use nfs as storage domain, but in your setup the HA is n= ot guaranteed: if a physical machine goes down and it's the one where the st= orage domain resides you are lost. Why not using gluster instead of nfs for t= he ovirt disks? You can still reserve a small gluster space for the non-ceph= machines (for example a cinder VM) and ceph for the rest. Where do you have= your cinder running?</div><div>Cheers,</div><div><br></div><div> &nbs= p; Alessandro</div><div><br>Il giorno 18 dic 2016, alle ore 18:05, rajatjpat= el <<a href=3D"mailto:rajatjpatel@gmail.com">rajatjpatel@gmail.com</a>>= ; ha scritto:<br><br></div><blockquote type=3D"cite"><div><div dir=3D"ltr"><= div class=3D"gmail_default" style=3D"font-family:comic sans ms,sans-serif;fo= nt-size:large;color:rgb(0,0,255)">Hi Alessandro,<br><br></div><div class=3D"= gmail_default" style=3D"font-family:comic sans ms,sans-serif;font-size:large= ;color:rgb(0,0,255)">Right now I have 2 physical server where I have host ov= irt these are HP proliant dl 380 each server 1*500GB SAS & 1TB *4 S= AS Disk and 1*500GB SSD. So right now I have use only one disk which 500GB o= f SAS for my ovirt to run on both server. rest are not in use. At present I a= m using NFS which coming from mapper to ovirt as storage, go forward we like= to use all these disk as hyper-converged for ovirt. RH I could see th= ere is KB for using gluster. But we are looking for Ceph bcoz best pref roma= nce and scale.<br><br><Screenshot from 2016-12-18 21-03-21.png><br></d= iv><div class=3D"gmail_default" style=3D"font-family:comic sans ms,sans-seri= f;font-size:large;color:rgb(0,0,255)">Regards<br></div><div class=3D"gmail_d= efault" style=3D"font-family:comic sans ms,sans-serif;font-size:large;color:= rgb(0,0,255)">Rajat<br></div></div><div class=3D"gmail_extra"><br clear=3D"a= ll"><div><div class=3D"gmail_signature" data-smartmail=3D"gmail_signature"><= div dir=3D"ltr"><div><font face=3D"tahoma, sans-serif" size=3D"4" style=3D"b= ackground-color:rgb(243,243,243)" color=3D"#0000ff">Hi</font></div><font fac= e=3D"tahoma, sans-serif" size=3D"4" style=3D"background-color:rgb(243,243,24= 3)" color=3D"#0000ff"><div><font face=3D"tahoma, sans-serif" size=3D"4" styl= e=3D"background-color:rgb(243,243,243)" color=3D"#0000ff"><br></font></div><= div><font face=3D"tahoma, sans-serif" size=3D"4" style=3D"background-color:r= gb(243,243,243)" color=3D"#0000ff"><br></font></div>Regards,<br>Rajat Patel<= br><br><a href=3D"http://studyhat.blogspot.com/" target=3D"_blank">http://st= udyhat.blogspot.com</a><br>FIRST THEY IGNORE YOU...<br>THEN THEY LAUGH AT YO= U...<br>THEN THEY FIGHT YOU...<br>THEN YOU WIN...</font><br><br></div></div>= </div> <br><div class=3D"gmail_quote">On Sun, Dec 18, 2016 at 8:49 PM, Alessandro D= e Salvo <span dir=3D"ltr"><<a href=3D"mailto:Alessandro.DeSalvo@roma1.inf= n.it" target=3D"_blank">Alessandro.DeSalvo@roma1.infn.it</a>></span> wrot= e:<br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-le= ft:1px #ccc solid;padding-left:1ex"><div dir=3D"auto"><div></div><div>Hi Raj= at,</div><div>sorry but I do not really have a clear picture of your actual s= etup, can you please explain a bit more?</div><div>In particular:</div><div>= <br></div><div>1) what to you mean by using 4TB for ovirt? In which machines= and how do you make it available to ovirt?</div><div><br></div><div>2) how d= o you plan to use ceph with ovirt?</div><div><br></div><div>I guess we can g= ive more help if you clarify those points.</div><div>Thanks,</div><div><br><= /div><div> Alessandro </div><div><div class=3D"h5"><div><br=
Il giorno 18 dic 2016, alle ore 17:33, rajatjpatel <<a href=3D"mailto:ra= jatjpatel@gmail.com" target=3D"_blank">rajatjpatel@gmail.com</a>> ha scri= tto:<br><br></div><blockquote type=3D"cite"><div><div dir=3D"ltr"><div><div>= <div><div><div>Great, thanks! Alessandro ++ Yaniv ++ <br><br></div>What I wa= nt to use around 4 TB of SAS disk for my Ovirt (which going to be RHV4.0.5 o= nce POC get 100% successful, in fact all product will be RH )<br><br></div>I= had done so much duckduckgo for all these solution and use lot of reference= from <a href=3D"http://ovirt.org" target=3D"_blank">ovirt.org</a> & <a h= ref=3D"http://access.redhat.com" target=3D"_blank">access.redhat.com</a> for= setting up a Ovirt engine and hyp.<br><br></div>We dont mind having more gu= est running and creating ceph block storage and which will be presented to o= virt as storage. Gluster is not is use right now bcoz we have DB will be run= ning on guest.<br><br></div>Regard<br></div>Rajat <br></div><br><div class=3D= "gmail_quote"><div dir=3D"ltr">On Sun, Dec 18, 2016 at 8:21 PM Alessandro De= Salvo <<a href=3D"mailto:Alessandro.DeSalvo@roma1.infn.it" target=3D"_bl= ank">Alessandro.DeSalvo@roma1.<wbr>infn.it</a>> wrote:<br></div><blockquo= te class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc sol= id;padding-left:1ex"><div dir=3D"auto" class=3D"m_324330166984056793gmail_ms= g"><div class=3D"m_324330166984056793gmail_msg"></div><div class=3D"m_324330= 166984056793gmail_msg">Hi,</div><div class=3D"m_324330166984056793gmail_msg"= having a 3-node ceph cluster is the bare minimum you can have to make it wo= rking, unless you want to have just a replica-2 mode, which is not safe.</di= v><div class=3D"m_324330166984056793gmail_msg">It's not true that ceph is no= t easy to configure, you might use very easily ceph-deploy, have puppet conf= iguring it or even run it in containers. Using docker is in fact the easiest= solution, it really requires 10 minutes to make a cluster up. I've tried it= both with jewel (official containers) and kraken (custom containers), and i= t works pretty well.</div><div class=3D"m_324330166984056793gmail_msg">The r= eal problem is not creating and configuring a ceph cluster, but using it fro= m ovirt, as it requires cinder, i.e. a minimal setup of openstack. We have i= t and it's working pretty well, but it requires some work. For your referenc= e we have cinder running on an ovirt VM using gluster.</div><div class=3D"m_= 324330166984056793gmail_msg">Cheers,</div><div class=3D"m_324330166984056793= gmail_msg"><br class=3D"m_324330166984056793gmail_msg"></div><div class=3D"m= _324330166984056793gmail_msg"> Alessandro </div></div><div d= ir=3D"auto" class=3D"m_324330166984056793gmail_msg"><div class=3D"m_32433016= 6984056793gmail_msg"><br class=3D"m_324330166984056793gmail_msg">Il giorno 1= 8 dic 2016, alle ore 17:07, Yaniv Kaul <<a href=3D"mailto:ykaul@redhat.co= m" class=3D"m_324330166984056793gmail_msg" target=3D"_blank">ykaul@redhat.co= m</a>> ha scritto:<br class=3D"m_324330166984056793gmail_msg"><br class=3D= "m_324330166984056793gmail_msg"></div><blockquote type=3D"cite" class=3D"m_3= 24330166984056793gmail_msg"><div class=3D"m_324330166984056793gmail_msg"><di= v dir=3D"ltr" class=3D"m_324330166984056793gmail_msg"><br class=3D"m_3243301= 66984056793gmail_msg"><div class=3D"gmail_extra m_324330166984056793gmail_ms= g"><br class=3D"m_324330166984056793gmail_msg"><div class=3D"gmail_quote m_3= 24330166984056793gmail_msg">On Sun, Dec 18, 2016 at 3:29 PM, rajatjpatel <sp= an dir=3D"ltr" class=3D"m_324330166984056793gmail_msg"><<a href=3D"mailto= :rajatjpatel@gmail.com" class=3D"m_324330166984056793gmail_msg" target=3D"_b= lank">rajatjpatel@gmail.com</a>></span> wrote:<br class=3D"m_324330166984= 056793gmail_msg"><blockquote class=3D"gmail_quote m_324330166984056793gmail_= msg" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"= <div dir=3D"ltr" class=3D"m_324330166984056793gmail_msg"><div class=3D"m_32= 4330166984056793m_-4293604042961126787m_-8155750194716306479gmail_signature m= _324330166984056793gmail_msg"><div dir=3D"ltr" class=3D"m_324330166984056793= gmail_msg"><div class=3D"m_324330166984056793gmail_msg">=E2=80=8BDear Team,<= br class=3D"m_324330166984056793gmail_msg"><br class=3D"m_324330166984056793= gmail_msg">We are using Ovirt 4.0 for POC what we are doing I want to check w= ith all Guru's Ovirt.<br class=3D"m_324330166984056793gmail_msg"><br class=3D= "m_324330166984056793gmail_msg">We have 2 hp proliant dl 380 with 500GB SAS &= amp; 1TB *4 SAS Disk and 500GB SSD.<br class=3D"m_324330166984056793gmail_ms= g"><br class=3D"m_324330166984056793gmail_msg">Waht we are done we have inst= all ovirt hyp on these h/w and we have physical server where we are running o= ur manager for ovirt. For ovirt hyp we are using only one 500GB of one HDD r= est we have kept for ceph, so we have 3 node as guest running on ovirt and f= or ceph. My question you all is what I am doing is right or wrong.<br class=3D= "m_324330166984056793gmail_msg"></div></div></div></div></blockquote><div cl= ass=3D"m_324330166984056793gmail_msg"><br class=3D"m_324330166984056793gmail= _msg"></div><div class=3D"m_324330166984056793gmail_msg">I think Ceph requir= es a lot more resources than above. It's also a bit more challenging to conf= igure. I would highly recommend a 3-node cluster with Gluster.</div><div cla= ss=3D"m_324330166984056793gmail_msg">Y.</div><div class=3D"m_324330166984056= 793gmail_msg"> </div><blockquote class=3D"gmail_quote m_324330166984056= 793gmail_msg" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-= left:1ex"><div dir=3D"ltr" class=3D"m_324330166984056793gmail_msg"><div clas= s=3D"m_324330166984056793m_-4293604042961126787m_-8155750194716306479gmail_s= ignature m_324330166984056793gmail_msg"><div dir=3D"ltr" class=3D"m_32433016= 6984056793gmail_msg"><div class=3D"m_324330166984056793gmail_msg"><br class=3D= "m_324330166984056793gmail_msg"></div><div class=3D"m_324330166984056793gmai= l_msg">Regards<br class=3D"m_324330166984056793gmail_msg"></div><div class=3D= "m_324330166984056793gmail_msg">Rajat=E2=80=8B</div><br class=3D"m_324330166= 984056793gmail_msg"></div></div> </div> <br class=3D"m_324330166984056793gmail_msg">______________________________<w= br>_________________<br class=3D"m_324330166984056793gmail_msg"> Users mailing list<br class=3D"m_324330166984056793gmail_msg"> <a href=3D"mailto:Users@ovirt.org" class=3D"m_324330166984056793gmail_msg" t= arget=3D"_blank">Users@ovirt.org</a><br class=3D"m_324330166984056793gmail_m= sg"> <a href=3D"http://lists.ovirt.org/mailman/listinfo/users" rel=3D"noreferrer"= class=3D"m_324330166984056793gmail_msg" target=3D"_blank">http://lists.ovir= t.org/<wbr>mailman/listinfo/users</a><br class=3D"m_324330166984056793gmail_= msg"> <br class=3D"m_324330166984056793gmail_msg"></blockquote></div><br class=3D"= m_324330166984056793gmail_msg"></div></div> </div></blockquote><blockquote type=3D"cite" class=3D"m_324330166984056793gm= ail_msg"><div class=3D"m_324330166984056793gmail_msg"><span class=3D"m_32433= 0166984056793gmail_msg">______________________________<wbr>_________________= </span><br class=3D"m_324330166984056793gmail_msg"><span class=3D"m_32433016= 6984056793gmail_msg">Users mailing list</span><br class=3D"m_324330166984056= 793gmail_msg"><span class=3D"m_324330166984056793gmail_msg"><a href=3D"mailt= o:Users@ovirt.org" class=3D"m_324330166984056793gmail_msg" target=3D"_blank"= Users@ovirt.org</a></span><br class=3D"m_324330166984056793gmail_msg"><span= class=3D"m_324330166984056793gmail_msg"><a href=3D"http://lists.ovirt.org/m= ailman/listinfo/users" class=3D"m_324330166984056793gmail_msg" target=3D"_bl= ank">http://lists.ovirt.org/<wbr>mailman/listinfo/users</a></span><br class=3D= "m_324330166984056793gmail_msg"></div></blockquote></div></blockquote></div>= <div dir=3D"ltr">-- <br></div><div data-smartmail=3D"gmail_signature"><p dir= =3D"ltr">Sent from my Cell Phone - excuse the typos & auto incorrect</p>=
</div> </div></blockquote></div></div></div></blockquote></div><br></div> </div></blockquote></body></html>= --Apple-Mail-F8806A8F-B8B6-48D9-BC6C-8A2672E65957--