[node-devel] ovirt-node CIM handling

Mike Burns mburns at redhat.com
Tue Apr 24 00:47:13 UTC 2012


Hi Geoff,

On Tue, 2012-04-24 at 07:49 +1000, Geoff O'Callaghan wrote:
> Hi All
> 
> 
> I notice that ovirt-node 2.3.0 had CIM handling added via
> bz 753215 / http://gerrit.ovirt.org/#change,2480 .   I thought i'd
> give it a test.  Alas, I can't get it to work.
> 
Yes, I failed to communicate that there were problems with the cim
stuff.  

The overall goal was to expose libvirt-cim schemas, not all cim schemas,
though that could be a next step.
> 
> I enable CIM via the TUI, set the password and attempt to query a well
> know CIMv2 class such as OMC_Processor - failed
> 
> 
> wbemcli ein -noverify
> https://cim:PASSWORD@10.100.0.3/root/cimv2:OMC_Processor
> *
> * wbemcli: Cim: (5) CIM_ERR_INVALID_CLASS: Class not found
> *
> 
> 
> So I thought i'd try enumerating all the classes.
> 
> 
> wbemcli ecn -noverify https://cim:passw0rd@10.100.0.3/root/cimv2
> 
> 
> returns nothing :-(
> 
> 
> and
> 
> 
> wbemcli ecn -noverify https://cim:PASSWORD@10.100.0.3/root/virt
> *
> * wbemcli: Cim: (3) CIM_ERR_INVALID_NAMESPACE: The target namespace
> does not exist
> *
> 

This is something that we caught that's related to the bugs you
reference later.  

> 
> this does imply at least that the cimv2 class exists :-)
> 
> 
> So, continuing the saga I went into the TUI, <F2> got myself a console
> and tried running sfcbd with all the debugging I could find.
> 
> 
> [root at kvm1 admin]# ps ax|grep sfcbd
>  2474 ?        Ssl    0:00 /usr/sbin/sfcbd -d
>  2475 ?        S      0:00 /usr/sbin/sfcbd -d
>  2479 ?        S      0:00 /usr/sbin/sfcbd -d
>  2480 ?        S      0:00 /usr/sbin/sfcbd -d
>  2579 ?        S      0:00 /usr/sbin/sfcbd -d
>  2580 ?        S      0:00 /usr/sbin/sfcbd -d
>  2584 ?        S      0:00 /usr/sbin/sfcbd -d
> 
> 
> I killed the current process and started as follows :
> 
> 
> [root at kvm1 admin]# SFCB_TRACE_FILE=/tmp/sfcb.log /usr/sbin/sfcbd -t
> 2097151 
> --- sfcbd V1.3.11 started - 4304
> --- Using /etc/sfcb/sfcb.cfg
> --- Max provider procs: 32
> --- initSocketPairs: 64
> --- localConnectServer started
> --- Max Http procs: 8
> --- sfcbd HTTP Daemon V1.3.11 configured for port 5989 - 4308
> --- sfcbd HTTP Daemon V1.3.11 configured for
> socket /tmp/sfcbHttpSocket - 4308
> --- Using Basic Authentication
> --- Keep-alive timeout: 15 seconds
> --- Maximum requests per connection: 10
> --- Caching ClassProvider
> for /var/lib/sfcb/registration/repository/root/interop/classSchemas
> (1.0-3) using 2296 bytes
> ^C--- Winding down sfcbd
> --- localConnectServer ended
> --- Stopping adapters
> --- HTTP-Daemon terminating 4308
> --- Adapters stopped
> --- Stopping providers
> ---  stopped IndCIMXMLHandler 4329
> ---  stopped InternalProvider 4317
> ---  stopped ProfileProvider 4314
> ---  stopped InteropProvider 4312
> ---  stopped ClassProvider 4309
> --- Providers stopped
> 
> 
> I can see the request come in via the logfile above
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(803) :
> --- Header: POST
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: Authorization: Basic blah
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: Host: 10.100.0.3:5989
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: Content-Type: application/xml; charset="utf-8"
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: Connection: Keep-Alive, TE
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: CIMProtocolVersion: 1.0
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: CIMOperation: MethodCall
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: CIMMethod: EnumerateClassNames
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: CIMObject: root%2Fcimv2
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: Content-Length: 398
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpAdapter.c(824) :
> --- Header: 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 ---
> httpAdapter.c(1000) : -#- xmlIn 398 bytes:
> <?xml version="1.0" encoding="utf-8" ?>
> <CIM CIMVERSION="2.0" DTDVERSION="2.0">
> <MESSAGE ID="4711" PROTOCOLVERSION="1.0"><SIMPLEREQ><IMETHODCALL
> NAME="EnumerateClassNames"><LOCALNAMESPACEPATH><NAMESPACE
> NAME="root"></NAMESPACE><NAMESPACE NAME="cimv2"></NAME
> SPACE></LOCALNAMESPACEPATH>
> <IPARAMVALUE NAME="DeepInheritance"><VALUE>TRUE</VALUE></IPARAMVALUE>
> </IMETHODCALL></SIMPLEREQ>
> </MESSAGE></CIM>
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 ---
> httpAdapter.c(1003) : -#- xmlIn end
> 
> 
> However the response is effectively null and that is how it's
> interpreted by the client
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(78) :
> "~PROTOCOLVERSION="1.0">\n<SIMPLERSP>\n<IMETHODRESPONSE~NAME="
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(79) : -<<
> xmlOut end
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(100) :
> Leaving: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(63) :
> Entering: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(69) : ->>
> xmlOut 19 bytes:
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(78) :
> EnumerateClassNames
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(79) : -<<
> xmlOut end
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(100) :
> Leaving: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(63) :
> Entering: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(69) : ->>
> xmlOut 18 bytes:
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(78) :
> ">\n<IRETURNVALUE>\n
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(79) : -<<
> xmlOut end
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(100) :
> Leaving: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(63) :
> Entering: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(69) : ->>
> xmlOut 0 bytes:
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(78) : 
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(79) : -<<
> xmlOut end
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(100) :
> Leaving: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(63) :
> Entering: commWrite
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(69) : ->>
> xmlOut 65 bytes:
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(78) :
> </IRETURNVALUE>\n</IMETHODRESPONSE>\n</SIMPLERSP>\n</MESSAGE>\n</CIM>
> 
> 
> [1] [04/23/2012 08:44:22] 4349/0x7f2be768e7c0 --- httpComm.c(79) : -<<
> xmlOut end
> 
> 
> 

Given the above output, you obviously know far more about cim than I do.
>From what I've learned, there is an incompatibility between sblim-sfcb
and libvirt-cim in Fedora 16.  We'll look closer as we move to F17 as
the base install.  
> 
> I tried this on F16 and centos6 - same result, so either i'm doing
> something terribly wrong or there is a bug in the upstream.
> 
> 
> Note:  I tried the 
> 
> 
> wbemcli ecn -noverify against an esxi box and got the expected result,
> so I can also confirm it's not a problem with my wbemcli client :)
> 

I've tested the same code on a RHEVH-6 machine and can connect correctly
to the libvirt-cim interface.
> 
> Just before I submitted this I found 
> 
> 
> https://bugzilla.redhat.com/show_bug.cgi?id=809090
> https://bugzilla.redhat.com/show_bug.cgi?id=809093
> 
> 
> I didn't find a bz against ovirt-node as well, should I raise one and xref with the above info and F16/17 bz's ?

You can, but there's not much we can do until libvirt-cim and sblim-sfcb
start working in Fedora.  Part of the problem is that there apparently
isn't much demand cim working in upstream.  

We're also looking for people that actually know something about cim to
help with developing and testing cim.

Thanks

Mike

> 
> Tks
> Geoff
> _______________________________________________
> node-devel mailing list
> node-devel at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/node-devel





More information about the node-devel mailing list