Skip to content

Split: the versioning, structure saving DAO

Utkarsh edited this page Jun 12, 2015 · 12 revisions

This document describes the split mongostore representation which

  • separates course structure from content where each course run (or partial course) can have its own structure and content can have many using structures
  • versions everything

It does not describe the original mongostore representation which combined structure and content and used the key to distinguish draft from published elements.

This document does not describe mongo nor its operations. See http://www.mongodb.org for information on Mongo.

Representation

This presentation on google docs provides an overview which may or may not be useful w/o narration.

The xmodule collections:

  • modulestore.active_versions: this collection maps the org, course, and run to the current draft and published versions of the course.
  • modulestore.structures: this collection has one entry per course run and one for the template.
  • modulestore.definitions: this collection has one entry per "module" or "block" version.

modulestore.active_versions: a simple map for dereferencing the correct course from the structures collection. It allows naming courses or structures, giving them addressable branches, and using them to fetch the "correct" version for any given course and branch. No course run will have more than one entry in this collection.

{
  "_id" : uniqueid,
  "versions" : { <versionName> : versionGuid, ...}
  "edited_by" : user_id,
  "edited_on" : date (native mongo rep)
}
  • _id is a unique id for finding this course run. It's a location-reference string, like mit.eng.eecs.6002x.industry.spring2013. (see Locators (successor to Locations))

  • versions: These are references to modulestore.structures. A location-reference like mit.eng.eecs.6002x.industry.spring2013/branch/draft refers to the value associated with draft for this document.

  • versionName is draft, published, or another user-defined string.

  • versionGuid is a system generated globally unique id (hash). It points to the entry in modulestore.structures

SplitMongo generates a new structure for each change to a structure: that is, for each move, deletion, node creation, or metadata change it adds a new structure and does not modify the old one. If the change was done using a pointer to a from active_index (i.e., course_id and branch), it will update the active_index entry to point to the new structure. No other pointers to the old structure will update. If the change is done using only a structure pointer or if Split determines that the head pointer has moved and the change is really occurring to an "old" structure, the head will not update but split will create a new structure (perhaps forking the structure's history).

Creating a course (creating a new run of a course or such) will create a new entry in this table with just a draft branch and will cause a copy of the corresponding entry in modulestore.structures. The entry in structures will point to its version parent in the source course.

modulestore.structures: the entries in this collection follow this definition:

{
  "_id" : course_guid,
  "blocks" : {
    block_guid : {  // the guid is an arbitrary id to represent this node in the course tree
      "fields" : {"children": [ block_guid* ], ...}
      "definition" : definition_guid,
      "category" : "section" | "sequence" | ...
      "edit_info" : {"edited_on" : date, "edited_by" : user, "previous_version": course_guid}
      ...// more guids
    },
  },
  "root" : block_guid,
  "original" : course_guid, // the first version of this course from which all others were derived
  "previous" : course_guid | null, // the previous revision of this course (null if this is the original)
  "edited_by" : user_id,
  "edited_on" : date
}
  • blocks: each block is a node in the course such as the course, a section, a subsection, a unit, or a component. The block ids remain the same over edits (they're not versioned).
  • root: the true top of the course. Not all nodes without parents are truly roots. Some are orphans or ancillary data (e.g., course_info).
  • course_guid, block_guid, definition_guid are not those specific strings but instead some system generated globally unique id.
    • The one which gets passed around and pointed to by urls is the block_guid; so, it will be the one the system ensures is readable. Unlike the other guids, this one stays the same over revisions and can even be the same between course runs (although the course run contextualizes it to distinguish its instantiated version).
  • definition points to the specific revision of the given element in modulestore.definitions which this version of the course includes.
  • children lists the block_guids which are the children of this node in the course tree. It's an error if the guid in the children list does not occur in the blocks dictionary. (This will change shortly to enable structures to include other structures by reference.)
  • the rest of fields will contain the other scope.settings (scope usage = 1, user = 0) fields.

modulestore.definitions: the data associated with each version of each node in the structures. Many courses may point to the same definition or may point to different versions derived from the same original definition.

{
  "_id" : guid,
  "fields" : ..,
  "default_settings" : {"display_name":..,..}, // a starting point for new uses of this definition
  "category" : xblocktype, // the xmodule/xblock type such as course, problem, html, video, about
  "edit_info" : {
    "original_version" : guid, // the first kept version of this definition from which all others were derived
    "previous_version" : guid | null, // the previous revision of this definition (null if this is the original)
    "edited_by" : user_id,  // the id of whomever pressed the draft or publish button
    "edited_on" : date
  }
}
  • _id: a guid to uniquely identify the definition.
  • fields is the json representation of the scope.content (definition = 1) fields used by the xmodule.
  • category is the xmodule type and used to figure out which xmodule to instantiate.

Templates

All field defaults will be defined through the xblock field.default mechanism. Templates, otoh, are for representing optional boilerplate usually for examples such as a multiple-choice problem or a video component with the fields all filled in. Templates are stored in yaml files which provide a template name, sorting and filtering information (e.g., requires advanced editor v allows simple editor), and then field: value pairs for setting xblocks' fields upon template selection.

Import/export

Export should allow the user to select the version of the course to export which can be any of the draft or published versions. At a minimum, the user should choose between draft or published.

Import should import the course as a draft course or whatever branch the user chooses regardless of whether it was exported as a published or draft one. If there's already a draft for the same course, in the best of all worlds, it would have the guid to see if the guid exists in the structures collection, and, if so, just make that the current structure (don't do any actual data changes). If there's no guid or the guid doesn't exist in the structures collection, then we'll need to work out the logic for how to decide what definitions to create v update v point to.

Location

The purpose of Locator is to identify content. That is, to be able to locate content by providing sufficient addressing. Our code needs to locate several types of things and uses several different types of locators for these. These are the types of things we need to address. Some of these can be the same as others, but I wanted to lay them out fairly fine grained here before proposing my distinctions:

  1. Courses: an object representing a course as an offering but not any of its content. Used for dashboards and other such navigators. These may specify a version or merely reference the idea of the course's existence.
  2. Course structures: the names (and other metadata), identifiers, and children pointers but not definitions for all the blocks in a course or a subtree of a course. Our applications often display contextual, outline, or other such structural information which do not need to include definitions but need to show display names, graded as, and other status info. This document's design makes fetching these a single document fetch; however, if it has to fetch the full course, it will require far more work (getting all definitions too) than the apps need.
  3. Blocks (uses of definitions within a version of a course including metadata, pointers to children, and type specific content)
  4. Definitions: use independent definitions of content without metadata (and currently w/o pointers to children).
  5. Version trees: Fetching the time history portrayal of a definition, course, or block including branching.
  6. Collections of courses, definitions, or blocks matching some partial descriptors (e.g., all courses for org x, all definitions of type foo, all blocks in course y of type x, all currently accessible courses (published with startdate < today and enddate > today)).
  7. Fetching of courses, blocks, or definitions via "human readable" urls. #6 (partial descriptors) may suffice for this as human readable does not guarantee uniqueness.

Some of these differ not so much in how to address them but in what should be returned. The content should be up to the functions not the addressing scheme. So, I think the addressable things are:

  1. Course as in #1 above: usually a specific offering of a course. Often used as a context for the other queries.
  2. Blocks (aka usages) as in #3 above: a specific block contextualized in a course
  3. Definitions (#4): a specific definition
  4. Collections of courses, blocks within a specific course, or definitions matching a partial descriptor</li

Course locator (course_loc)

There are 3 ways to locate a course:

  1. By its unique id in the active_versions collection with an implied or specified selection of draft or published version.
  2. By its org and prettyid in the active_versions collection with an implied or specified selection of draft or published version. The server should raise an error on any attempt to use this addressing if it does not uniquely identify an entry in the active_versions collection. (For now, I'm only supporting this in the partial locator)
  3. By its unique id in the structures collection.

Block locator (block_loc)

A block locator finds a specific node in a specific version of a course. Thus, it needs a course locator plus a block_guid.

Definition locator (definition_loc)

Just a guid.

Partial descriptor collections locators (partial)

In the most general case, and to simplify implementation, these can be any payload passable to mongo for doing the lookup. The specification of which collection to look into can be implied by which lookup function your code calls (get_courses, get_blocks, get_definitions) or we could add it as another property. For now, I will leave this as merely a search string. Thus, to find all courses for org = mitx, {"org": "mitx"}. To find all blocks in a course whose display name contains "circuit example", call get_blocks with the course locator plus {"fields.display_name" : /circuit example/i} (the i makes it case insensitive and is just an example). To find if a definition is used in a course, call get_blocks with the course locator plus {definition : definition_guid}. Note, this looks for a specific version of the definition. If you wanted to see if it used any of a set of versions, use {definition : {"$in" : [definition_guid*]}}

Publishing

Publishing is the process of making content which is available on one branch also available on another branch. Usually the source branch is a 'draft' or 'editing' branch and the destination branch is a 'published' or 'live' branch. Sometimes we may want the destination branches to be preview, staging, alpha, beta, A-test, honors, or other variants. The new representation makes arbitrary branches very easy.

The difficulty in publishing is that a common use case is to only publish some parts of the course. For example, publish weeks 1 - 3 but not the quiz in week 3 because it's still being edited. Publishing by itself does not imply that the students can see the material because the LMS uses start dates and other mechanisms to decide whether content should be accessible. The reason that partial publication is a "problem" is that it cannot use the simple expediency of just pointing the destination branch to the same structure version as the source branch. Instead it needs to create a new version snapshot which is a subset of the source version snapshot using the same identities and definitions.

In the old Mongo, there were only 2 branches: live and draft. All course content above verticals (all nodes 2 or more levels above the leaves) was in both at the same time. The act of publishing was the act of copying verticals or leaf components from 'draft' to live. All changes to elements above verticals immediately impacted both the draft and live course (element crud, moving, attribute setting). Live courses constantly stumble over pointers to 'private' verticals generating ItemNotFoundErrors and slowing down some pages significantly by repeatedly searching for the private verticals. These will not be problems in split mongo.

In the new Modulestore, the authoring team can select how many branches to support. They could have one: the live one is the draft one. All changes take effect immediately. They could have dozens (draft, experimental, alpha, staged, honors, live, scaffolded, ...). They could publish from any of these to any of the others although in practice that wouldn't make sense.

Publishing a node from one branch to another must imply:

  • ensure all of that node's ancestors are in the destination branch. If not, throw an exception.
  • put the node is in the destination branch in the same relative order vis-a-vis its published siblings under its parent. NOTE: to ensure the resulting order is determinate, publishing may involve changing the order of previously published siblings. For example, if publishing [c, d] and source parent has children [a, b, c, d, e] and destination parent currently has children [e, b], there's no obviously correct resulting order; thus, publish will reorder destination to [b, c, d, e] to make it conformant with the destination.
  • ensure the fields of the node in the destination are the same as the source

What's subject to opinion and how the design will progress:

  • publish the node's children to the destination unless they were explicitly excluded via a blacklist provided to the publish api. (That is, publishing works on subtrees not individual nodes).
  • if the previously published node had children which no longer exist in the source, remove those from the destination (unless they are also in the blacklist).
  • if the node being published is deleted in the source but exists in the destination, then the semantics are to delete the node from the destination.
  • any node deletion implies deleting the whole subtree with that given root.

Common publishing use cases

  • Update a given node's fields (e.g., display name, dates, grading policy, or change the order of subnodes)
  • Move a node from one parent to another
  • Publish a "ready" subtree except for some specific not quite ready subtrees under it

Publish API

publish( course_locator_w_source_branch, destination_branch, subtree_list, blacklist )

  • course_locator_w_source_branch: a CourseLocator with either a specific version id or a branch from which publishing will occur.
  • destination_branch: can be either a branch id (just a string) or a CourseLocator with a branch id. Cannot be a version id as the act of publishing implies updating the head of a branch? It is possible to publish from one course to another; however, the block ids may conflict (the old branch's Chapter4 may be overwritten to something which has no similarity nor commonality). In this case, the system won't do anything to ensure the resulting tree is truly a tree versus a dag where a node may occur more than once.
  • subtree_list: subtree root block ids where the root and all of its descendants except those in blacklist will be copied from source to destination. Can be simply block ids or Locators. If Locators, it's an error if any reference any structure other than the one referenced by course_locator_w_source_branch.
  • blacklist: the subtree roots not to copy from source to destination. Follows the same rules as subtree_list and directly subtracts subtrees from subtree_list. Note, if any of these nodes already exist in destination, they will remain in their current state. To delete, you need to include the parent in subtree_list and not put the node in blacklist thus implying making the node the same in destination as in source (deleted).

All changes will occur as if happening in one transaction. Even if the changes to destination_branch result in numerous versions, the head won't update until the system commits all changes. Thus, the LMS users' experience should maintain consistency.

In addition to all explicitly listed nodes, publish will ensure each subtree root's parent exists in destination. It will raise exceptions if any parents are missing.

Any exceptions will cause publish to fail rather than to be partial.

Clone this wiki locally