- Users - oVirt List Archives

Disabling migration when there is a problem with storage
by Mark Steele 03 Feb '16

03 Feb '16

Hello, I'd like to know if there is a way to configure ovirt (oVirt Engine Version: 3.5.0.1-1.el6) to NOT migrate VM's when there is a problem with attached storage. Scenario - our storage array that hosts all our VM disks failed. When that happened, ovirt attempted to migrate running VM's from one host to another. Unfortunately this left many of the VM's in a 'migrating to' state that could only be resolved by restarting the HV's I'd prefer to have ovirt simply hang the VM's until disk IO returns. Is this possible? *** *Mark Steele* CIO / VP Technical Operations | TelVue Corporation TelVue - We Share Your Vision 800.885.8886 x128 | msteele(a)telvue.com | http://www.telvue.com twitter: http://twitter.com/telvue | facebook: https://www.facebook.com/telvue

3 2

HE multipath problem.
by gflwqs gflwqs 03 Feb '16

03 Feb '16

Hi List, I have installed the hosted engine over iscsi but have a problem getting it to install over the multipath, even though i can see the multipath device from the host with multipath -ll. When i get to the point in the HE installer where i should choose the disk i only see the underlying devices like sdb sdc. So if i choose one of those it results in that it will only use one of the paths! My questions is: - Is there a bug in the installer script or am i doing something wrong? - I now have a running HE and would like to manually change the single path to dual, how do i accomplish this? I am running ovirt 3.5.6.2. Regards Christian

3 6

increase file descriptor
by alireza sadeh seighalan 03 Feb '16

03 Feb '16

hi everyone this is my server information: OS: CentOS7.1 (latest update) server: DL 380 G7 ovirt: 3.6.2 [root@hv ~]# cat /proc/sys/fs/file-max 4896581 [root@hv ~]# cat /proc/sys/fs/file-nr 10416 0 4896581 [root@hv ~]# ulimit -a core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) unlimited pending signals (-i) 192428 max locked memory (kbytes, -l) 64 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 real-time priority (-r) 0 stack size (kbytes, -s) 8192 cpu time (seconds, -t) unlimited max user processes (-u) 192428 virtual memory (kbytes, -v) unlimited file locks (-x) unlimited my open files seem to be increased: [root@hv ~]# lsof |wc -l 1454641 acording my information , do i need increase file descriptor? how can i increase them? thanks inadvance

1 0

Re: [ovirt-users] [ovirt-devel] Hello and A Question about oVirt
by Michal Skrivanek 03 Feb '16

03 Feb '16

> On 03 Feb 2016, at 06:36, zhukaijie <kjzhu14(a)is.ac.cn> wrote: > > > ________________________________________ > 发件人: Michal Skrivanek [mskrivan(a)redhat.com] > 发送时间: 2016年2月2日 17:55 > 收件人: zhukaijie > 抄送: devel(a)ovirt.org > 主题: Re: [ovirt-devel] Hello and A Question about oVirt > > On 02 Feb 2016, at 10:40, Yaniv Dary <ydary(a)redhat.com<mailto:ydary@redhat.com>> wrote: > > I don't think we have a option like this. Michal? > > > Yaniv Dary > Technical Product Manager > Red Hat Israel Ltd. > 34 Jerusalem Road > Building A, 4th floor > Ra'anana, Israel 4350109 > > Tel : +972 (9) 7692306 > 8272306 > Email: ydary(a)redhat.com<mailto:ydary@redhat.com> > IRC : ydary > > On Mon, Feb 1, 2016 at 5:16 AM, zhukaijie <kjzhu14(a)is.ac.cn<mailto:kjzhu14@is.ac.cn>> wrote: > Hello, now I have defined a custom property named 'A' in oVirt Engine. Administrator is responsible for entering the value (and arbitrary string ) of 'A' before starting the VM. After an users trys to start the VM in oVirt, VDSM will add the value of 'A' in the qemu:arg of libvirt domain xml, so that the value of 'A' will be added into the QEMU Cmd as a param. However, just like the password of VNC or SPICE, I want to hide the value of 'A' in '*' format in both Libvirt domain xml and QEMU Cmd, So could you please tell me how to achieve it? Thank you very much and happy 2016. > > No, I don’t think you would be able to make libvirt and qemu to hide it. Unfortunately it would be exposed…for log files you are protected by file access permissions, but if there is anything sensitive on the command line and you have a user who can get a shell on that machine one can always see that in process listing > > do you perhaps need to pass some secret to a VM? Might be better via payload, it can be accessed in the guest as a file then. > > Thanks, > michal > > _______________________________________________ > Devel mailing list > Devel(a)ovirt.org<mailto:Devel@ovirt.org> > http://lists.ovirt.org/mailman/listinfo/devel > > Thank you. But there is still a doubt for me. In vdsm/graphics.py, function _setPasswd uses "*****" format to hide the true password of VNC and SPICE if disableticketing feature is not used. So later how can Libvirt translates the "*****" format into true password? Thank you. for password field it’s an exception and it’s explicitly logged with *. of course the proper secret password is supplied to libvirt. But as a generic field elsewhere …they are not getting hidden….all the parameters would look like ***** which is not helpful:)

1 0

importing existing storage domain
by Charles Kozler 02 Feb '16

02 Feb '16

Say I have backed up a storage domain to another NFS. I now want to restore that backup. What would be the best method to restore all the disk images on that domain? I have tried export domains and importing existing domains but every time it says they are not empty and they "may be from somewhere else". They are in fact from somewhere else but that is somewhat the point of the exercise I am attempting to simulate a restore-from-backup (basically just an rsync of the entire storage domain to an offsite location) and its not going as planned. I have all the VM images and can ensure they are all in tact but ovirt refuses to acknowledge it I have also tried setting up a new storage domain and then copying the backup images directory to the new images directory under the newly established storage domain. ovirt sees the used space go down but does not reference / acknowledge / view any of the data actually there unless it creates it itself How can I restore an rsync backup of a storage domain? I am feeling like there isnt a way to readily do it so if not, how can I restore these backed up images? -- *Charles Kozler* *Vice President, IT Operations* FIX Flyer, LLC 225 Broadway | Suite 1600 | New York, NY 10007 1-888-349-3593 http://www.fixflyer.com <http://fixflyer.com> NOTICE TO RECIPIENT: THIS E-MAIL IS MEANT ONLY FOR THE INTENDED RECIPIENT(S) OF THE TRANSMISSION, AND CONTAINS CONFIDENTIAL INFORMATION WHICH IS PROPRIETARY TO FIX FLYER LLC. ANY UNAUTHORIZED USE, COPYING, DISTRIBUTION, OR DISSEMINATION IS STRICTLY PROHIBITED. ALL RIGHTS TO THIS INFORMATION IS RESERVED BY FIX FLYER LLC. IF YOU ARE NOT THE INTENDED RECIPIENT, PLEASE CONTACT THE SENDER BY REPLY E-MAIL AND PLEASE DELETE THIS E-MAIL FROM YOUR SYSTEM AND DESTROY ANY COPIES.

4 6

Migrate Hosted Engine to new iSCSI san with new IP/Target
by bill＠taney.com 02 Feb '16

02 Feb '16

--Apple-Mail=_4564AA9F-2C27-414A-915B-0C6E64D8F21D Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 In a hosted-engine deployment has anyone had to migrate to a new iSCSI = SAN? I can not figure out how to move the engine(s) to a new iSCSI = target.. The SAN=E2=80=99s can fully replicate to one another and I can = move the LUN to the other device but don=E2=80=99t see how to stop the = engine and change the settings for it=E2=80=99s iSCSI target to another = IP/IQN. Any direction would be greatly appreciated - I am looking for = any method to get this done, we have to retire the old SAN - Thanks.=EF=BB= =BF= --Apple-Mail=_4564AA9F-2C27-414A-915B-0C6E64D8F21D Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8 <html><body style=3D"word-wrap: break-word; -webkit-nbsp-mode: space; = -webkit-line-break: after-white-space;" class=3D""><span = style=3D"background-color: rgb(255, 255, 255);" class=3D""><font = color=3D"#404040" size=3D"2" class=3D"">In a hosted-engine deployment = has anyone had to migrate to a new iSCSI SAN? I can not figure out how = to move the engine(s) to a new iSCSI target.. The SAN=E2=80=99s can = fully replicate to one another and I can move the LUN to the other = device but don=E2=80=99t see how to stop the engine and change the = settings for it=E2=80=99s iSCSI target to another IP/IQN. Any direction = would be greatly appreciated - I am looking for any method to get this = done, we have to retire the old SAN - = Thanks.=EF=BB=BF</font></span></body></html>= --Apple-Mail=_4564AA9F-2C27-414A-915B-0C6E64D8F21D--

4 13

deleted host still showing up in hosted engine vm-status
by Charles Kozler 02 Feb '16

02 Feb '16

Using a hosted engine on 3.6 and I deleted a host I recently added but the system is still showing up in hosted-engine --vm-status How can I ensure this system is fully deleted? -- *Charles Kozler* *Vice President, IT Operations* FIX Flyer, LLC 225 Broadway | Suite 1600 | New York, NY 10007 1-888-349-3593 http://www.fixflyer.com <http://fixflyer.com> NOTICE TO RECIPIENT: THIS E-MAIL IS MEANT ONLY FOR THE INTENDED RECIPIENT(S) OF THE TRANSMISSION, AND CONTAINS CONFIDENTIAL INFORMATION WHICH IS PROPRIETARY TO FIX FLYER LLC. ANY UNAUTHORIZED USE, COPYING, DISTRIBUTION, OR DISSEMINATION IS STRICTLY PROHIBITED. ALL RIGHTS TO THIS INFORMATION IS RESERVED BY FIX FLYER LLC. IF YOU ARE NOT THE INTENDED RECIPIENT, PLEASE CONTACT THE SENDER BY REPLY E-MAIL AND PLEASE DELETE THIS E-MAIL FROM YOUR SYSTEM AND DESTROY ANY COPIES.

2 1

Re: [ovirt-users] [3.6] Delete node
by Douglas Schilling Landgraf 02 Feb '16

02 Feb '16

This is a multi-part message in MIME format. --------------040303000609030907090806 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Hi Mario, On 02/01/2016 02:12 PM, [Syci] Mario Montes wrote: > Hi, yes I put on maintenance first, but the option is disable. The > node is fail. I need delete force. Looks like your host is in fail not in maintenance, you can force it with the below commands and later remove it. step#1 On Engine machine: $ su - postgres $ psql -d engine engine# update vds_dynamic set status=2 where vds_id in (select vds_id from vds_static where vds_name LIKE '%THE_NODE_HOST_NAME%'); step#2 Go back to admin portal, your host will be in maint. and you should be able to remove it. Please keep the list in the reply for other people might help and provide the logs from engine/vdsm. > > ------------------------------------------------------------------------ > *De: *"Douglas Schilling Landgraf" <dougsland(a)redhat.com> > *Para: *"[Syci] Mario Montes" <mario.montes(a)syci.es>, "users" > <users(a)ovirt.org> > *Enviados: *Lunes, 1 de Febrero 2016 19:07:23 > *Asunto: *Re: [ovirt-users] [3.6] Delete node > > Hi, > > On 02/01/2016 01:54 PM, [Syci] Mario Montes wrote: > > Hello, I need delete one node, but the option "Delete" is not > active. How I do it? > > > Have you tried to put the host into maintenance first? > http://www.ovirt.org/OVirt_Administration_Guide#Removing_a_Host > > --------------040303000609030907090806 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 7bit <html> <head> <meta content="text/html; charset=utf-8" http-equiv="Content-Type"> </head> <body bgcolor="#FFFFFF" text="#000000"> Hi Mario,<br> <br> <div class="moz-cite-prefix">On 02/01/2016 02:12 PM, [Syci] Mario Montes wrote:<br> </div> <blockquote cite="mid:2109534935.261948.1454353978570.JavaMail.zimbra@syci.es" type="cite"> <div style="font-family: tahoma, new york, times, serif; font-size: 10pt; color: #000000"> <div>Hi, yes I put on maintenance first, but the option is disable. The node is fail. I need delete force. <br> </div> </div> </blockquote> Looks like your host is in fail not in maintenance, you can force it with the below commands and later remove it.<br> <br> step#1<br> On Engine machine:<br> <br> $ su - postgres<br> $ psql -d engine<br> engine# update vds_dynamic set status=2 where vds_id in (select vds_id from vds_static where vds_name LIKE '%THE_NODE_HOST_NAME%');<br> <br> step#2<br> Go back to admin portal, your host will be in maint. and you should be able to remove it.<br> <br> Please keep the list in the reply for other people might help and provide the logs from engine/vdsm.<br> <br> <blockquote cite="mid:2109534935.261948.1454353978570.JavaMail.zimbra@syci.es" type="cite"> <div style="font-family: tahoma, new york, times, serif; font-size: 10pt; color: #000000"> <div><br> </div> <hr id="zwchr" data-marker="__DIVIDER__"> <div data-marker="__HEADERS__"><b>De: </b>"Douglas Schilling Landgraf" <a class="moz-txt-link-rfc2396E" href="mailto:dougsland@redhat.com"><dougsland(a)redhat.com></a><br> <b>Para: </b>"[Syci] Mario Montes" <a class="moz-txt-link-rfc2396E" href="mailto:mario.montes@syci.es"><mario.montes(a)syci.es></a>, "users" <a class="moz-txt-link-rfc2396E" href="mailto:users@ovirt.org"><users(a)ovirt.org></a><br> <b>Enviados: </b>Lunes, 1 de Febrero 2016 19:07:23<br> <b>Asunto: </b>Re: [ovirt-users] [3.6] Delete node<br> </div> <br> <div data-marker="__QUOTED_TEXT__">Hi,<br> <br> <div class="moz-cite-prefix">On 02/01/2016 01:54 PM, [Syci] Mario Montes wrote:<br> </div> <blockquote cite="mid:1399555782.261898.1454352867918.JavaMail.zimbra@syci.es"> <div style="font-family: tahoma, new york, times, serif; font-size: 10pt; color: #000000"> <div>Hello, I need delete one node, but the option "Delete" is not active. How I do it?</div> </div> </blockquote> <br> Have you tried to put the host into maintenance first? <br> <a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://www.ovirt.org/OVirt_Administration_Guide#Removing_a_Host" target="_blank">http://www.ovirt.org/OVirt_Administration_Guide#Removing_a_Host</a><br> <br> <br> </div> </div> </blockquote> <br> </body> </html> --------------040303000609030907090806--

2 1

Disk copies
by Pavel Gashev 02 Feb '16

02 Feb '16

--_000_1454233087326010camelacroniscom_ Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: base64 SGVsbG8NCg0KSSBqdXN0IHRyaWVkIHRvIGF0dGFjaCBhIGNvcHkgb2YgYSBkaXNrIChtYWRlIGZv ciBiYWNrdXAgcHVycG9zZXMpIHRvIGEgVk0uIEV2ZXJ5dGhpbmcgd29ya3Mgd2VsbCBleGNlcHQg bGl2ZSBzbmFwc2hvdCBkZWxldGlvbiBkb2Vzbid0IHdvcmsgYW55bW9yZS4gVGhlIGlzc3VlIGlz IHRoYXQgdGhlIGRpc2sgaGFzIHZvbHR5cGU9U0hBUkVELCBpbnN0ZWFkIG9mIExFQUYuIExpdmUg c25hcHNob3QgbWVyZ2luZyBkb2Vzbid0IHdvcmsgb24gc2hhcmVkIGRpc2tzLg0KDQpDb3VsZCBz b21lb25lIGNsYXJpZnksIHdoeSBkb2VzIGRpc2sgY29weWluZyBtYWtlIGRpc2tzIHNoYXJlZD8g SXMgaXQgYSBidWcsIG9yIHRoZXJlIGlzIGEgaGlkZGVuIHJlYXNvbiBmb3IgdGhhdD8NCg0KVGhh bmtzDQoNCg== --_000_1454233087326010camelacroniscom_ Content-Type: text/html; charset="utf-8" Content-ID: <D890FBDB9441EB4FA7F9B9F2265DA432(a)acronis.com> Content-Transfer-Encoding: base64 PGh0bWw+DQo8aGVhZD4NCjxtZXRhIGh0dHAtZXF1aXY9IkNvbnRlbnQtVHlwZSIgY29udGVudD0i dGV4dC9odG1sOyBjaGFyc2V0PXV0Zi04Ij4NCjwvaGVhZD4NCjxib2R5Pg0KPGRpdj5IZWxsbzwv ZGl2Pg0KPGRpdj48YnI+DQo8L2Rpdj4NCjxkaXY+SSBqdXN0IHRyaWVkIHRvIGF0dGFjaCBhIGNv cHkgb2YgYSBkaXNrIChtYWRlIGZvciBiYWNrdXAgcHVycG9zZXMpIHRvIGEgVk0uIEV2ZXJ5dGhp bmcgd29ya3Mgd2VsbCBleGNlcHQgbGl2ZSBzbmFwc2hvdCBkZWxldGlvbiBkb2Vzbid0IHdvcmsg YW55bW9yZS4gVGhlIGlzc3VlIGlzIHRoYXQgdGhlIGRpc2sgaGFzIHZvbHR5cGU9U0hBUkVELCBp bnN0ZWFkIG9mIExFQUYuIExpdmUgc25hcHNob3QgbWVyZ2luZyBkb2Vzbid0IHdvcmsgb24NCiBz aGFyZWQgZGlza3MuPC9kaXY+DQo8ZGl2Pjxicj4NCjwvZGl2Pg0KPGRpdj5Db3VsZCBzb21lb25l IGNsYXJpZnksIHdoeSBkb2VzIGRpc2sgY29weWluZyBtYWtlIGRpc2tzIHNoYXJlZD8gSXMgaXQg YSBidWcsIG9yIHRoZXJlIGlzIGEgaGlkZGVuIHJlYXNvbiBmb3IgdGhhdD88L2Rpdj4NCjxkaXY+ PGJyPg0KPC9kaXY+DQo8ZGl2PlRoYW5rczwvZGl2Pg0KPGRpdj48YnI+DQo8L2Rpdj4NCjwvYm9k eT4NCjwvaHRtbD4NCg== --_000_1454233087326010camelacroniscom_--

2 7

oVirt hook to manage physical CPU pool, and automatically pin CPUs without overlap on the host
by David Jorm 02 Feb '16

02 Feb '16

Hi All Thanks for the help that people have recently provided on this list. I'm running into another edge case that, as far as I can see, does not have a native solution within oVirt. I have a use case where every VM must have 4 pinned vCPUs, and these must map 1:1 to physical CPUs so that there are never two VMs contending for the same physical cores. I already have code in place that ensures a VM will not be created or started if n/4 VMs are already running on the target host, where n is the number of physical cores. What I need now is to ensure that pinning is managed so that each physical core maps 1:1 to vCPUs. Would it be possible to manage the pool of physical cores in a local DB, and implement a VDSM hook similar to the example on this page? http://www.ovirt.org/VDSM-Hooks Or would there be a preferable solution? Thanks David

2 3