% % Copyright (c) 2006-2007 XenSource, Inc. % % Permission is granted to copy, distribute and/or modify this document under % the terms of the GNU Free Documentation License, Version 1.2 or any later % version published by the Free Software Foundation; with no Invariant % Sections, no Front-Cover Texts and no Back-Cover Texts. A copy of the % license is included in the section entitled % "GNU Free Documentation License" or the file fdl.tex. % % Authors: Ewan Mellor, Richard Sharp, Dave Scott, Jon Harrop. % The API is presented here as a set of Remote Procedure Calls, with a wire format based upon XML-RPC. No specific language bindings are prescribed, although examples will be given in the python programming language. Although we adopt some terminology from object-oriented programming, future client language bindings may or may not be object oriented. The API reference uses the terminology {\em classes\/} and {\em objects\/}. For our purposes a {\em class\/} is simply a hierarchical namespace; an {\em object\/} is an instance of a class with its fields set to specific values. Objects are persistent and exist on the server-side. Clients may obtain opaque references to these server-side objects and then access their fields via get/set RPCs. %In each class there is a $\mathit{uid}$ field that assigns an indentifier %to each object. This $\mathit{uid}$ serves as an object reference %on both client- and server-side, and is often included as an argument in %RPC messages. For each class we specify a list of fields along with their {\em types\/} and {\em qualifiers\/}. A qualifier is one of: \begin{itemize} \item $\mathit{RO}_\mathit{run}$: the field is Read Only. Furthermore, its value is automatically computed at runtime. For example: current CPU load and disk IO throughput. \item $\mathit{RO}_\mathit{ins}$: the field must be manually set when a new object is created, but is then Read Only for the duration of the object's life. For example, the maximum memory addressable by a guest is set before the guest boots. \item $\mathit{RW}$: the field is Read/Write. For example, the name of a VM. \end{itemize} A full list of types is given in Chapter~\ref{api-reference}. However, there are three types that require explicit mention: \begin{itemize} \item $t~\mathit{Ref}$: signifies a reference to an object of type $t$. \item $t~\mathit{Set}$: signifies a set containing values of type $t$. \item $(t_1, t_2)~\mathit{Map}$: signifies a mapping from values of type $t_1$ to values of type $t_2$. \end{itemize} Note that there are a number of cases where {\em Ref}s are {\em doubly linked\/}---e.g.\ a VM has a field called {\tt VIFs} of type $(\mathit{VIF}~\mathit{Ref})~\mathit{Set}$; this field lists the network interfaces attached to a particular VM. Similarly, the VIF class has a field called {\tt VM} of type $(\mathit{VM}~{\mathit Ref})$ which references the VM to which the interface is connected. These two fields are {\em bound together\/}, in the sense that creating a new VIF causes the {\tt VIFs} field of the corresponding VM object to be updated automatically. The API reference explicitly lists the fields that are bound together in this way. It also contains a diagram that shows relationships between classes. In this diagram an edge signifies the existence of a pair of fields that are bound together, using standard crows-foot notation to signify the type of relationship (e.g.\ one-many, many-many). \section{RPCs associated with fields} Each field, {\tt f}, has an RPC accessor associated with it that returns {\tt f}'s value: \begin{itemize} \item ``{\tt get\_f(Ref x)}'': takes a {\tt Ref} that refers to an object and returns the value of {\tt f}. \end{itemize} Each field, {\tt f}, with attribute {\em RW} and whose outermost type is {\em Set\/} has the following additional RPCs associated with it: \begin{itemize} \item an ``{\tt add\_to\_f(Ref x, v)}'' RPC adds a new element v to the set\footnote{ % Since sets cannot contain duplicate values this operation has no action in the case that {\tt v} was already in the set. % }; \item a ``{\tt remove\_from\_f(Ref x, v)}'' RPC removes element {\tt v} from the set; \end{itemize} Each field, {\tt f}, with attribute {\em RW} and whose outermost type is {\em Map\/} has the following additional RPCs associated with it: \begin{itemize} \item an ``{\tt add\_to\_f(Ref x, k, v)}'' RPC adds new pair {\tt (k, v)} to the mapping stored in {\tt f} in object {\tt x}. Adding a new pair for duplicate key, {\tt k}, overwrites any previous mapping for {\tt k}. \item a ``{\tt remove\_from\_f(Ref x, k)}'' RPC removes the pair with key {\tt k} from the mapping stored in {\tt f} in object {\tt x}. \end{itemize} Each field whose outermost type is neither {\em Set\/} nor {\em Map\/}, but whose attribute is {\em RW} has an RPC acessor associated with it that sets its value: \begin{itemize} \item For {\em RW\/} ({\em R\/}ead/{\em W\/}rite), a ``{\tt set\_f(Ref x, v)}'' RPC function is also provided. This sets field {\tt f} on object {\tt x} to value {\tt v}. \end{itemize} \section{RPCs associated with classes} \begin{itemize} \item Each class has a {\em constructor\/} RPC named ``{\tt create}'' that takes as parameters all fields marked {\em RW\/} and $\mathit{RO}_\mathit{ins}$. The result of this RPC is that a new {\em persistent\/} object is created on the server-side with the specified field values. \item Each class has a {\tt get\_by\_uuid(uuid)} RPC that returns the object of that class that has the specified {\tt uuid}. \item Each class that has a {\tt name\_label} field has a ``{\tt get\_by\_name\_label(name)}'' RPC that returns a set of objects of that class that have the specified {\tt label}. \item Each class has a ``{\tt destroy(Ref x)}'' RPC that explicitly deletes the persistent object specified by {\tt x} from the system. This is a non-cascading delete -- if the object being removed is referenced by another object then the {\tt destroy} call will fail. \end{itemize} \subsection{Additional RPCs} As well as the RPCs enumerated above, some classes have additional RPCs associated with them. For example, the {\tt VM} class has RPCs for cloning, suspending, starting etc. Such additional RPCs are described explicitly in the API reference.