UPDATE: I finally cracked it. Even though the systemvm.iso files were updated into the xen servers, there was a .vhd file that is created into the local storage of the servers that are running the system VMs that contains the systemvm.iso files. Therefore, I had to stop the CS MSs delete those .vhd files from the local storage of my xenservers and everything was good to go.
Thanks for your support Min ;) On Thu, May 29, 2014 at 10:03 AM, Rafael Weingartner < rafaelweingart...@gmail.com> wrote: > Hi, > Before I ran any kind of command I decided to dive a little deeper into > this issue. > So, I checked the systemvm.iso that is inside CS MS. and it turned out > that it has the jars for the CS agent 4.3.0. > However, I also checked the systemvm.iso that is in the xen hosts in > "/usr/share/xcp/packages/iso/". That .iso also has the jars for the 4.3.0 > agent. So, I do not get why my systems VMs have the jar for the 4.1.1 > version. > > Can somebody explain me the process, like when the jars that are in the > systemvm.iso are copied to the system vm? which scripts do that? is there > any way to debug that process? I do not understand how the jars from 4.1.1 > version are in the Systemvm of 4.3.0 version. > > > > On Wed, May 28, 2014 at 6:16 PM, Min Chen <min.c...@citrix.com> wrote: > >> I don't know the exact cmd line to clear that, and I have always used >> XenCenter UI to do that, where clear tags in General tabs. >> >> Thanks >> -min >> >> On 5/28/14 2:11 PM, "Rafael Weingartner" <rafaelweingart...@gmail.com> >> wrote: >> >> >by clear, you mean "xe vdi-destroy uuid=vdi of systemvm.iso"? >> >So, I just need to stop the management servers, run that in every cluster >> >that I have and everything should be fine? >> > >> > >> >On Wed, May 28, 2014 at 5:42 PM, Min Chen <min.c...@citrix.com> wrote: >> > >> >> You should clear old systemvm.iso from xen server so that new >> >>systemvm.iso >> >> built can be deployed there. >> >> >> >> Thanks >> >> -min >> >> >> >> On 5/28/14 12:42 PM, "Rafael Weingartner" <rafaelweingart...@gmail.com >> > >> >> wrote: >> >> >> >> >Another interesting fact: >> >> >All the jars, cloud-**.jar are label as if they were 4.1.1 version. >> >> >Is that right? >> >> > >> >> > >> >> >On Wed, May 28, 2014 at 4:01 PM, Rafael Weingartner < >> >> >rafaelweingart...@gmail.com> wrote: >> >> > >> >> >> Just an update. >> >> >> I have just checked my console proxy VM, which is also with "--" in >> >> >>agent >> >> >> state at "insfrastructure>System VM" page. >> >> >> The agente is running there, however its state is "--". >> >> >> >> >> >> I also looked into the source code, and the class " >> >> >> org.apache.cloudstack.storage.resource.NfsSecondaryStorageResource" >> >>is >> >> >> located in a jar called "cloud-secondary-storage" which is not in >> the >> >> >> directory "/usr/local/cloud/systemvm" that contains the jars used to >> >>run >> >> >> the agent. Neither in my SSVM nor in Console proxy VM that jar >> exist. >> >> >> >> >> >> I do not get, I cannot be the only one with this problem. It seems >> >>that >> >> >> this jar is missing into the system vm template. >> >> >> >> >> >> >> >> >> On Wed, May 28, 2014 at 2:45 PM, Rafael Weingartner < >> >> >> rafaelweingart...@gmail.com> wrote: >> >> >> >> >> >>> thanks, I checked the "/var/log/cloud/cloud.out", And I found this >> >> >>>error: >> >> >>> ERROR AgentShell:607 - Unable to start agent: Resource class not >> >> >>>found: >> >> >>> org.apache.cloudstack.storage.resource.NfsSecondaryStorageResource >> >>due >> >> >>>to: >> >> >>> java.lang.ClassNotFoundException: >> >> >>> org.apache.cloudstack.storage.resource.NfsSecondaryStorageResource >> >> >>> Unable to start agent: Resource class not found: >> >> >>> org.apache.cloudstack.storage.resource.NfsSecondaryStorageResource >> >>due >> >> >>>to: >> >> >>> java.lang.ClassNotFoundException: >> >> >>> org.apache.cloudstack.storage.resource.NfsSecondaryStorageResource >> >> >>> >> >> >>> It seems that there is a jar dependency missing. >> >> >>> >> >> >>> >> >> >>> >> >> >>> On Wed, May 28, 2014 at 2:31 PM, Min Chen <min.c...@citrix.com> >> >>wrote: >> >> >>> >> >> >>>> Your UI indicates that agent is not running on your SSVM, it >> should >> >> >>>>show >> >> >>>> as 'Running', that is why your cloud.host does not have an entry >> >>for >> >> >>>>your >> >> >>>> new SSVM, thus we cannot find a remote host to send command to. >> >>Agent >> >> >>>>is >> >> >>>> a >> >> >>>> java code running inside your SSVM, you should see what error you >> >> >>>> encountered in starting agent in your ssvm from cloud.out file in >> >>SSVM >> >> >>>> log. >> >> >>>> >> >> >>>> Thanks >> >> >>>> -min >> >> >>>> >> >> >>>> >> >> >>>> On 5/28/14 10:20 AM, "Rafael Weingartner" >> >> >>>><rafaelweingart...@gmail.com> >> >> >>>> wrote: >> >> >>>> >> >> >>>> >What agent are you talking about? >> >> >>>> >Is it an app that run into the SSVM? >> >> >>>> >I am using xenserver and XCP as the hypervisor. >> >> >>>> > >> >> >>>> >I looked into the CS UI, infrastructure>System VMs and the column >> >> >>>>that >> >> >>>> has >> >> >>>> >some values about the agent-state has a -- in it. Is it normal? >> >> >>>> > >> >> >>>> >At the end, should the SSVM have an entry into the host table? >> >> >>>> > >> >> >>>> > >> >> >>>> >On Wed, May 28, 2014 at 2:05 PM, Min Chen <min.c...@citrix.com> >> >> >>>>wrote: >> >> >>>> > >> >> >>>> >> Check if agent is running your ssvm. >> >> >>>> >> >> >> >>>> >> Thanks >> >> >>>> >> -min >> >> >>>> >> >> >> >>>> >> On 5/28/14 7:26 AM, "Rafael Weingartner" >> >> >>>><rafaelweingart...@gmail.com >> >> >>>> > >> >> >>>> >> wrote: >> >> >>>> >> >> >> >>>> >> >Hi all, >> >> >>>> >> >sorry, to bother you again with this thread. >> >> >>>> >> >Could at least someone that has the CS 4.3.0 deployed try to >> >> >>>> download a >> >> >>>> >> >volume and/or a template? >> >> >>>> >> > >> >> >>>> >> >Could someone also check the database, table ³cloud.host² and >> >> >>>>look if >> >> >>>> >> >there >> >> >>>> >> >is an entry there for the running SSVM and console proxy VM? >> >> >>>> >> > >> >> >>>> >> > >> >> >>>> >> > >> >> >>>> >> >On Mon, May 26, 2014 at 4:42 PM, Rafael Weingartner < >> >> >>>> >> >rafaelweingart...@gmail.com> wrote: >> >> >>>> >> > >> >> >>>> >> >> Hi folks, I am sending this to the dev list as well, hence I >> >>am >> >> >>>>not >> >> >>>> >>sure >> >> >>>> >> >> if it is or not a bug. Sorry if I am mistaken. >> >> >>>> >> >> >> >> >>>> >> >> So, after I upgraded CS from 4.1.1 to 4.3.0, everything >> >>seemed >> >> >>>>to >> >> >>>> be >> >> >>>> >> >> working just fine, however, whenever I try to download a >> >> >>>>template >> >> >>>> or >> >> >>>> >> >>volume >> >> >>>> >> >> I am getting the following error: >> >> >>>> >> >> ERROR [o.a.c.s.d.d.CloudStackImageStoreDriverImpl] >> >> >>>> >> >> (Job-Executor-1:ctx-1cb86faa ctx-61f7e4f9) No remote >> >>endpoint to >> >> >>>> send >> >> >>>> >> >> command, check if host or ssvm is down? >> >> >>>> >> >> ERROR [o.a.c.s.d.d.CloudStackImageStoreDriverImpl] >> >> >>>> >> >> (Job-Executor-1:ctx-1cb86faa ctx-61f7e4f9) Unable to create >> a >> >> >>>>link >> >> >>>> >>for >> >> >>>> >> >> entity at ... on ssvm,No remote endpoint to send command, >> >>check >> >> >>>>if >> >> >>>> >>host >> >> >>>> >> >>or >> >> >>>> >> >> ssvm is down >> >> >>>> >> >> >> >> >>>> >> >> I have already destroyed the system VMs in order to let CS >> >> >>>>recreate >> >> >>>> >>them >> >> >>>> >> >> with the new system VM template. I logged into the SSVM to >> >>check >> >> >>>> >> >>whether or >> >> >>>> >> >> not it was created using the new template and it indeed. >> >> >>>> Therefore, I >> >> >>>> >> >>have >> >> >>>> >> >> ruled out that problem. >> >> >>>> >> >> >> >> >>>> >> >> Then, I took a look into the source code, more specifically >> >>at >> >> >>>> >> >> >> >> >>>> >> >> >> >>>> >> >> >>>> >> >> >> >> >>>>>>>>>>"org.apache.cloudstack.storage.datastore.driver.CloudStackImageSt >> >>>>>>>>>>or >> >> >>>>>>>>eDri >> >> >>>> >>>>ve >> >> >>>> >> >>rImpl.createEntityExtractUrl(DataStore, >> >> >>>> >> >> String, ImageFormat, DataObject)" >> >> >>>> >> >> and >> >> >>>> >> >> >> >> >>>> >> >> >> >> >>>> >> >> >> >>>> >> >> >>>> >> >> >> >> >>>>>>>>>>"org.apache.cloudstack.storage.endpoint.DefaultEndPointSelector.f >> >>>>>>>>>>in >> >> >>>>>>>>dEnd >> >> >>>> >>>>po >> >> >>>> >> >>intForImageStorage(DataStore)". >> >> >>>> >> >> >> >> >>>> >> >> I noticed that it looks for an entry at table "Cloud.host" >> >>that >> >> >>>>has >> >> >>>> >>the >> >> >>>> >> >> type ³SecondaryStorageVM². >> >> >>>> >> >> >> >> >>>> >> >> I took a look into the DB, but, there are no entries either >> >>for >> >> >>>> SSVM >> >> >>>> >>or >> >> >>>> >> >> Console proxy in ³Cloud.host² table. I mean no entries for >> >>the >> >> >>>>VMs >> >> >>>> >>that >> >> >>>> >> >>are >> >> >>>> >> >> running. The older VMs that were destroyed have an entry >> >>there. >> >> >>>> >> >> >> >> >>>> >> >> Have anyone seem a problem like this before? >> >> >>>> >> >> Is there something I am missing? >> >> >>>> >> >> >> >> >>>> >> >> The VMs are running, should not them be registered there? >> >> >>>> >> >> >> >> >>>> >> >> -- >> >> >>>> >> >> Rafael Weingärtner >> >> >>>> >> >> >> >> >>>> >> > >> >> >>>> >> > >> >> >>>> >> > >> >> >>>> >> >-- >> >> >>>> >> >Rafael Weingärtner >> >> >>>> >> >> >> >>>> >> >> >> >>>> > >> >> >>>> > >> >> >>>> >-- >> >> >>>> >Rafael Weingärtner >> >> >>>> >> >> >>>> >> >> >>> >> >> >>> >> >> >>> -- >> >> >>> Rafael Weingärtner >> >> >>> >> >> >> >> >> >> >> >> >> >> >> >> -- >> >> >> Rafael Weingärtner >> >> >> >> >> > >> >> > >> >> > >> >> >-- >> >> >Rafael Weingärtner >> >> >> >> >> > >> > >> >-- >> >Rafael Weingärtner >> >> > > > -- > Rafael Weingärtner > -- Rafael Weingärtner
