Xen project Mailing List

[Xen-devel] [RFC]Some questions about xen-4.1.1

To: xen-users@xxxxxxxxxxxxx, xen-devel@xxxxxxxxxxxxxxxxxxx

From: harryxiyou <harryxiyou@xxxxxxxxx>

Date: Sun, 26 Feb 2012 01:39:01 +0800

Authentication-results: mr.google.com; spf=pass (google.com: domain of harryxiyou@xxxxxxxxx designates 10.112.103.228 as permitted sender) smtp.mail=harryxiyou@xxxxxxxxx; dkim=pass header.i=harryxiyou@xxxxxxxxx

Cc: cloudxy@xxxxxxxxxxxxxxxx, Kang Hua <kanghua151@xxxxxxxxx>, çåå <kangqi1988@xxxxxxxxx>

Delivery-date: Sat, 25 Feb 2012 17:39:51 +0000

List-id: Xen developer discussion <xen-devel.lists.xen.org>

Hi list, Maybe i should not ask questions here. If true, please forgive me ;-) I build xen environment under Centos-5.4.My env is like this. [2012-02-25 23:51:19 4789] INFO (XendDomainInfo:3269) Mounting /home/jiawei/workshop1/1.sys.img on /dev/xvdp. [2012-02-25 23:51:19 4789] DEBUG (DevController:95) DevController: writing {'backend-id': '0', 'virtual-device': '51952', 'device-type': 'disk', 'state': '1', 'backend': '/local/domain/0/backend/vbd/0/51952'} to /local/domain/0/device/vbd/51952. Linux local00212201021a.zte.com.cn 2.6.32.39 #1 SMP Sun Feb 19 11:21:09 CST 2012 x86_64 x86_64 x86_64 GNU/Linux # xm info host : local00212201021a.zte.com.cn release : 2.6.32.39 version : #1 SMP Sun Feb 19 11:21:09 CST 2012 machine : x86_64 nr_cpus : 2 nr_nodes : 1 cores_per_socket : 2 threads_per_core : 1 cpu_mhz : 1795 hw_caps : bfebfbff:20100800:00000000:00000940:0000e31d:00000000:00000001:00000000 virt_caps : total_memory : 984 free_memory : 665 free_cpus : 0 xen_major : 4 xen_minor : 1 xen_extra : .1 xen_caps : xen-3.0-x86_64 xen-3.0-x86_32p xen_scheduler : credit xen_pagesize : 4096 platform_params : virt_start=0xffff800000000000 xen_changeset : unavailable xen_commandline : cc_compiler : gcc version 4.1.2 20080704 (Red Hat 4.1.2-51) cc_compile_by : root cc_compile_domain : zte.com.cn cc_compile_date : Thu Feb 23 23:32:01 CST 2012 xend_config_format : 4 ================================ [a] We get an error by running following steps. ================================ 1, Making a delta plate by vhd-util # vhd-util snapshot -n /home/jiawei/workshop1/1.sys.img -p /home/jiawei/workshop1/domU-x86_64-FS.img -m 2ïstart a domu # xm cr pygrub.conf Using config file "./pygrub.conf". Error: Disk isn't accessible # cat pygrub.conf name = "ttylinux" memory = 512 vcpus = 1 vif = [ 'bridge=virbr0' ] on_reboot = 'restart' on_crash = 'restart' bootloader = "/usr/bin/pygrub" disk = ['tap2:vhd:/home/jiawei/workshop1/1.img,sda,w'] Be sure we have a correct raw image /home/jiawei/workshop1/domU-x86_64-FS.img becasue we have tested it. But in this condition we can get a tapdisk device. Details are here. [root@local00212201021a workshop1]# tap-ctl create -a vhd:/home/jiawei/workshop1/1.img /dev/xen/blktap-2/tapdev0 [root@local00212201021a workshop1]# tap-ctl list 5835 0 0 vhd /home/jiawei/workshop1/1.img At last we take some xen logs from directory /var/log/xen/xen.log like this. ... [2012-02-25 23:51:19 4789] DEBUG (XendDomainInfo:103) XendDomainInfo.create(['vm', ['name', 'ttylinux'], ['memory', 512], ['on_reboot', 'restart'], ['on_crash', 'restart'], ['on_xend_start', 'ignore'], ['on_xend_stop', 'ignore'], ['vcpus', 1], ['oos', 1], ['bootloader', '/usr/bin/pygrub'], ['bootloader_args', '-q'], ['image', ['linux', ['videoram', 4], ['tsc_mode', 0], ['nomigrate', 0]]], ['s3_integrity', 1], ['device', ['tap2', ['uname', 'tap2:vhd:/home/jiawei/workshop1/1.sys.img'], ['dev', 'xvda'], ['mode', 'w']]], ['device', ['vif', ['bridge', 'virbr0']]]]) [2012-02-25 23:51:19 4789] DEBUG (XendDomainInfo:2498) XendDomainInfo.constructDomain [2012-02-25 23:51:19 4789] DEBUG (balloon:187) Balloon: 541836 KiB free; need 16384; done. [2012-02-25 23:51:19 4789] DEBUG (XendDomain:476) Adding Domain: 34 [2012-02-25 23:51:19 4789] DEBUG (XendDomainInfo:2836) XendDomainInfo.initDomain: 34 256 [2012-02-25 23:51:19 4789] INFO (XendDomainInfo:3269) Mounting /home/jiawei/workshop1/1.sys.img on /dev/xvdp. [2012-02-25 23:51:19 4789] DEBUG (DevController:95) DevController: writing {'backend-id': '0', 'virtual-device': '51952', 'device-type': 'disk', 'state': '1', 'backend': '/local/domain/0/backend/vbd/0/51952'} to /local/domain/0/device/vbd/51952. [2012-02-25 23:51:19 4789] DEBUG (DevController:97) DevController: writing {'domain': 'Domain-0', 'frontend': '/local/domain/0/device/vbd/51952', 'uuid': '6772f76d-11b0-246e-b483-e9fc1056df13', 'bootable': '0', 'dev': '/dev/xvdp', 'state': '1', 'params': '/dev/xen/blktap-2/tapdev1', 'mode': 'r', 'online': '1', 'frontend-id': '0', 'type': 'phy', 'tapdisk-params': 'vhd:/home/jiawei/workshop1/1.sys.img'} to /local/domain/0/backend/vbd/0/51952. [2012-02-25 23:51:19 4789] DEBUG (DevController:144) Waiting for 51952. [2012-02-25 23:51:19 4789] DEBUG (DevController:628) hotplugStatusCallback /local/domain/0/backend/vbd/0/51952/hotplug-status. [2012-02-25 23:51:20 4789] DEBUG (DevController:628) hotplugStatusCallback /local/domain/0/backend/vbd/0/51952/hotplug-status. [2012-02-25 23:51:20 4789] DEBUG (DevController:642) hotplugStatusCallback 1. [2012-02-25 23:51:20 4789] DEBUG (DevController:144) Waiting for 51952. [2012-02-25 23:51:20 4789] DEBUG (DevController:628) hotplugStatusCallback /local/domain/0/backend/vbd/0/51952/hotplug-status. [2012-02-25 23:51:20 4789] DEBUG (DevController:642) hotplugStatusCallback 1. [2012-02-25 23:51:20 4789] ERROR (XendBootloader:43) Disk isn't accessible ... We think that there are something wrong with the device /dev/xvdp. And its *DevController* process 51952 can not get its corrective status for hotplug. So error happens to us. Right? ================================================= [b]But we can launch a domu like following steps successfully ================================================= 1, Tap-ctl create a device # tap-ctl create -a vhd:/home/jiawei/workshop1/1.img /dev/xen/blktap-2/tapdev0 # tap-ctl list 5835 0 0 vhd /home/jiawei/workshop1/1.img 2, Start domu # xm cr pygrub1.conf Using config file "./pygrub1.conf". Started domain ttylinux (id=44) # cat pygrub1.conf memory = 512 name = "ttylinux" bootloader = "/usr/bin/pygrub" disk = ['phy:/dev/xen/blktap-2/tapdev0,xvda1,w'] By up steps, we do get a domu. Could anyone tell me the detail differences between [a] and [b]? Why does [a] lead to an error? And [b] is always right for us? =========================== [c] Some other confusions =========================== Sometimes i can not get a device by 'tap-ctl create -a args'. It says i happen to some errors. After i do separate steps of 'tap-ctl create', it looks well for me. I will show you a concrete example. the wrong way ------------------------ # tap-ctl create -a tap2:vhd:/home/jiawei/workshop1/1.sys.img # tap-ctl list # tap-ctl create -a tapdisk:vhd:/home/jiawei/workshop1/1.sys.img # cat /var/log/messages | tail -20 ... Feb 25 22:50:20 local00212201021a tapdisk2[1171]: I/O queue driver: lio Feb 25 22:50:20 local00212201021a tapdisk2[1171]: received 'attach' message (uuid = 0) Feb 25 22:50:20 local00212201021a tapdisk2[1171]: sending 'attach response' message (uuid = 0) Feb 25 22:50:20 local00212201021a tapdisk2[1171]: received 'open' message (uuid = 0) Feb 25 22:50:20 local00212201021a tapdisk2[1171]: sending 'error' message (uuid = 0) Feb 25 22:50:20 local00212201021a tap-ctl: tap-err:tap_ctl_open: open failed, err -2 Feb 25 22:50:20 local00212201021a tapdisk2[1171]: received 'detach' message (uuid = 0) Feb 25 22:50:20 local00212201021a tapdisk2[1171]: sending 'detach response' message (uuid = 0) ..... the right way -------------------- [root@local00212201021a workshop1]# tap-ctl attach usage: attach <-p pid> <-m minor> [root@local00212201021a workshop1]# tap-ctl list 1194 - - - - [root@local00212201021a workshop1]# tap-ctl attach -p 1194 -m 0 [root@local00212201021a workshop1]# tap-ctl list 1194 - - - - [root@local00212201021a workshop1]# tap-ctl allocate /dev/xen/blktap-2/tapdev0 [root@local00212201021a workshop1]# tap-ctl list 1194 - - - - - 0 - - - [root@local00212201021a workshop1]# tap-ctl attach -p 1194 -m 0 [root@local00212201021a workshop1]# tap-ctl list 1194 0 0 - - [root@local00212201021a workshop1]# tap-ctl open usage: open <-p pid> <-m minor> <-a args> [root@local00212201021a workshop1]# tap-ctl open -p 1194 -m 0 -a vhd:/home/jiawei/workshop1/1.sys.img [root@local00212201021a workshop1]# tap-ctl list 1194 0 0 vhd /home/jiawei/workshop1/1.sys.img After 'the wrong way' and 'the right way', i think the two ways have different mechanisms. But i don't understand them clearly. Cloud anyone tell me the ones ? -- Thanks Harry Wei _______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxx http://lists.xen.org/xen-devel

©2013 Xen Project, A Linux Foundation Collaborative Project. All Rights Reserved.
Linux Foundation is a registered trademark of The Linux Foundation.
Xen Project is a trademark of The Linux Foundation.