Difference between revisions of "Trove-Replication-And-Clustering-API-Single"
(→Show Instance) |
(Adding some notes from a design perspective) |
||
(2 intermediate revisions by one other user not shown) | |||
Line 151: | Line 151: | ||
{ | { | ||
"id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", | "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", | ||
− | "name": "product-a" | + | "name": "product-a" |
− | |||
− | |||
− | |||
}, | }, | ||
{ | { | ||
Line 160: | Line 157: | ||
"name": "product-b", | "name": "product-b", | ||
"mysql": { | "mysql": { | ||
− | |||
"slave_of": [{"id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998"}], | "slave_of": [{"id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998"}], | ||
"read_only": true | "read_only": true | ||
Line 170: | Line 166: | ||
</pre> | </pre> | ||
<br> | <br> | ||
+ | |||
==== Remove Replication (aka "Promote" to Standalone) ==== | ==== Remove Replication (aka "Promote" to Standalone) ==== | ||
<br> | <br> | ||
Line 212: | Line 209: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products", |
"join": false | "join": false | ||
} | } | ||
Line 233: | Line 230: | ||
<br> | <br> | ||
Notes: | Notes: | ||
− | * Enforce ' | + | * Enforce 'replSet' field to be provided for MongoDB, even in the case of a standalone/single instance. See http://www.mongodb.com/blog/post/dont-let-your-standalone-mongodb-server-stand-alone for reasoning. |
* 'join' indicates whether you're joining an existing replica-set, or creating a new one. If 'join' is false, and an active replica-set by that name for the tenant already exists, the request will be failed. 'join' by default will be false, but was included above for illustrative purposes. | * 'join' indicates whether you're joining an existing replica-set, or creating a new one. If 'join' is false, and an active replica-set by that name for the tenant already exists, the request will be failed. 'join' by default will be false, but was included above for illustrative purposes. | ||
* 'type' is 'member' vs. 'primary' because in a replica-set, the primary is dynamic and can change in an election. | * 'type' is 'member' vs. 'primary' because in a replica-set, the primary is dynamic and can change in an election. | ||
Line 250: | Line 247: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products", |
"join": true | "join": true | ||
} | } | ||
Line 271: | Line 268: | ||
<br> | <br> | ||
Notes: | Notes: | ||
− | * If 'join' is true, and there is no existing replica-set for the tenant matching the ' | + | * If 'join' is true, and there is no existing replica-set for the tenant matching the 'replSet' value, the request will be failed. |
* Will have to use 'db.isMaster()' to determine the current primary to execute replica-set commands against (since it can be dynamic due to elections) | * Will have to use 'db.isMaster()' to determine the current primary to execute replica-set commands against (since it can be dynamic due to elections) | ||
* Will use http://docs.mongodb.org/manual/tutorial/expand-replica-set/#configure-and-add-a-member | * Will use http://docs.mongodb.org/manual/tutorial/expand-replica-set/#configure-and-add-a-member | ||
Line 291: | Line 288: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products", |
"join": true | "join": true | ||
} | } | ||
Line 328: | Line 325: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
}, | }, | ||
Line 356: | Line 353: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
}, | }, | ||
Line 365: | Line 362: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
}, | }, | ||
Line 374: | Line 371: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
} | } | ||
Line 395: | Line 392: | ||
"mongodb": { | "mongodb": { | ||
"type": "arbiter", | "type": "arbiter", | ||
− | " | + | "replSet": "products", |
"join": true | "join": true | ||
} | } | ||
Line 429: | Line 426: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products", |
"join": true, | "join": true, | ||
"priority": 0, | "priority": 0, | ||
Line 454: | Line 451: | ||
<br> | <br> | ||
Notes: | Notes: | ||
− | * 'type', ' | + | * 'type', 'replSet', 'join', 'priority', 'hidden', and 'slaveDelay' are the only fields supported in topology.mongodb{}. All other configuration values must be set via a configuration-group. After more thought, consider supporting 'hostname' and 'votes' as well. |
* Why isn't 'priority', 'hidden' and 'slaveDelay' in a configuration-group you ask? This is explained in "Modifying a Replica-Set" below. | * Why isn't 'priority', 'hidden' and 'slaveDelay' in a configuration-group you ask? This is explained in "Modifying a Replica-Set" below. | ||
<br> | <br> | ||
Line 523: | Line 520: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
}, | }, | ||
Line 532: | Line 529: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
}, | }, | ||
Line 541: | Line 538: | ||
"mongodb": { | "mongodb": { | ||
"type": "member", | "type": "member", | ||
− | " | + | "replSet": "products" |
} | } | ||
} | } | ||
Line 554: | Line 551: | ||
* It should now be clear why 'priority', 'hidden' and 'slaveDelay' are in topology.mongodb{} vs. a configuration-group: when a configuration-group is changed, an event is immediately triggered to update any attached trove instances. Therefore, if you have a heterogeneous mixture of configuration-groups in a replica-set, there is no way to coordinate a consolidated rs.reconfig(). | * It should now be clear why 'priority', 'hidden' and 'slaveDelay' are in topology.mongodb{} vs. a configuration-group: when a configuration-group is changed, an event is immediately triggered to update any attached trove instances. Therefore, if you have a heterogeneous mixture of configuration-groups in a replica-set, there is no way to coordinate a consolidated rs.reconfig(). | ||
* Downside: if topology.mongodb{} may have fields returned on a GET that you cannot change in a PATCH/PUT; re-worded, the granularity of what is permissible to change in a PATCH becomes complicated to check and validate. | * Downside: if topology.mongodb{} may have fields returned on a GET that you cannot change in a PATCH/PUT; re-worded, the granularity of what is permissible to change in a PATCH becomes complicated to check and validate. | ||
− | * TBD on what should be returned in the mongodb{} on a GET /instance/:id and GET /instance/:id/topology. It's a question of whether we should persist anything beyond the 'type' and ' | + | * TBD on what should be returned in the mongodb{} on a GET /instance/:id and GET /instance/:id/topology. It's a question of whether we should persist anything beyond the 'type' and 'replSet'. If say the 'priority' is stored, you introduce the possibility of drift from the truth, but can easily return it on a GET; if it's not stored, do we prompt MongoDB for the truth on a GET, or is that too computationally expensive? |
<br> | <br> | ||
<br> | <br> | ||
Line 1,443: | Line 1,440: | ||
<br> | <br> | ||
Work in Progress | Work in Progress | ||
+ | |||
+ | == Feedback == | ||
+ | |||
+ | ==== Masters/Slaves ==== | ||
+ | *drewford: Can one make a master/slave out of an existing instance? | ||
+ | *drewford: Type and version should follow the same schema as listing datastore types | ||
+ | |||
+ | ==== Arbiters ==== | ||
+ | *drewford: Could/should arbiters be managed by the system automatically? Is it worth putting that responsibility on the API user, or can it be automated by the system for an easier, less technical experience? From reading the article noted in the link, it seems like it's a good enough idea to automate it to reduce the need for users to research and decide for themselves. | ||
+ | |||
+ | ==== Delayed Members ==== | ||
+ | *drewford: Could/should delayed members be managed by the system automatically? Can it just be fully automated for a better, less technical experience? | ||
+ | |||
+ | ==== Topology ==== | ||
+ | *drewford: Members of the topology in the mongo example all have the same "mongodb" object. Unless this could also be "mysql" or "redis" in the same topology, it might make sense to move up "type" and "replSet" attributes as siblings to "name". | ||
+ | *drewford: The "type" attribute also seems redundant to the name of the array. Are there attributes other than "members" on a "topology"? If not, can "topology" be the array with each object having a "type"? | ||
+ | |||
+ | ==== Cassandra Create Cluster ==== | ||
+ | *drewford: can "num_tokens", "endpoint_snitch" or "auto_bootstrap" be automated in any way so they receive a valid default value even if the user does not specify one? Looking for ways in which we can take the decision workload off the user, and put it on the system in a healthy way. | ||
+ | |||
+ | ==== Promote/Disconnect Redis Slave ==== | ||
+ | *drewford: Seems like the promotion/disconnection json should be structured the same across all datastores if possible - with the backend handling any special manipulation to get the job done. Standardize in favor of the user. | ||
+ | |||
+ | ==== Redis Cluster - Create Cluster ==== | ||
+ | *drewford: "redis_cluster" here is labelled differently than the same object in Cassandra and Couchbase - they don't use "_cluster". |
Latest revision as of 19:16, 12 May 2014
MySQL Master/Slave
Create Master
Request:
POST /instances { "instance": { "availability_zone": "us-west-1", "name": "product-a", "datastore": { "type": "mysql", "version": "mysql-5.5" }, "configuration": "b9c8a3f8-7ace-4aea-9908-7b555586d7b6", "flavorRef": "7", "volume": { "size": 1 } } }
Response:
{ "instance": { "status": "BUILD", "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", "configuration": { "id": "b9c8a3f8-7ace-4aea-9908-7b555586d7b6", "name": "config-a", "links": [{...}] }, ... } }
Create Slave
Request:
POST /instances { "instance": { "availability_zone": "us-west-2", "name": "product-b", "datastore": { "type": "mysql", "version": "mysql-5.5" }, "topology": { "mysql": { "slave_of": [{"id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998"}] "read_only": true } }, "configuration": "fc318e00-3a6f-4f93-af99-146b44912188", "flavorRef": "7",, "volume": { "size": 1 } } }
Response:
{ "instance": { "status": "BUILD", "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", "configuration": { "id": "fc318e00-3a6f-4f93-af99-146b44912188", "name": "config-b", "links": [{...}] }, ... } }
Notes:
- For master/slave wirings, the 'server_id' must differ between master and slave, and optionally the slave can specify whether it is read-only or not to avoid accidental writes.
- Update: Agreed on 03/14/14 that the user should not have to specify 'server_id'. Instead, trove will be responsible for setting it and ensuring that a slave does not have the same server_id as its master, or any sibling slaves. As a part of this agreement, this requires removing 'server_id' from configuration-groups overrides (to avoid the user meddling with our bookkeeping).
- 'read_only' was removed from configuration-groups and moved to topology.mysql{} because forcing the user to create a configuration-group for every MySQL slave is arduous and a poor user experience.
- 'slave_of' and 'read_only' will be the only supported input fields for topology.mysql{}
- 'read_only' is only permitted if 'slave_of' is set, otherwise the request will be failed.
- Opinion: 'read_only' should default to true
- 'slave_of' is an array to properly represent multi-source replication in the future (coming in MySQL 5.7)
- For now, 'slave_of' requires a vanilla trove instance uuid, but will inevitably need to be prefixed with namespacing to support multiple dcs and sources (e.g. trove:us-west:tenant_id:instance:dfbbd9ca-b5e1-4028-adb7-f78643e17998)
- Why topology.<datastore>{} vs. topology.metadata{}? This should be more evident as you walk through the rest of the datatores, but in short: it provides strong(er) typing, allows for apischema validation, easier to document, simpler to maintain, etc.
Show Instance
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998
Response:
{ "instance": { "status": "ACTIVE", "updated": "2014-02-16T03:38:49" "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", "datastore": { "version": "mysql-5.5", "type": "mysql", }, "flavor": { "id": "7", "links": [{...}] }, "configuration": { "id": "b9c8a3f8-7ace-4aea-9908-7b555586d7b6", "name": "config-a", "links": [{...}] } } }
Show Topology
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology
Response:
{ "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a" }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", "mysql": { "slave_of": [{"id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998"}], "read_only": true } } ] } }
Remove Replication (aka "Promote" to Standalone)
Request:
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "mysql": { "promote": { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859" } } }
Response:
TBD
MongoDB
Create Replica-Set
Request:
POST /instances { "instance": { "name": "product-a", ... "datastore": { "type": "mongodb", "version": "mongodb-2.0.4" }, "topology": { "mongodb": { "type": "member", "replSet": "products", "join": false } }, ... } }
Response:
{ "instance": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", ... } }
Notes:
- Enforce 'replSet' field to be provided for MongoDB, even in the case of a standalone/single instance. See http://www.mongodb.com/blog/post/dont-let-your-standalone-mongodb-server-stand-alone for reasoning.
- 'join' indicates whether you're joining an existing replica-set, or creating a new one. If 'join' is false, and an active replica-set by that name for the tenant already exists, the request will be failed. 'join' by default will be false, but was included above for illustrative purposes.
- 'type' is 'member' vs. 'primary' because in a replica-set, the primary is dynamic and can change in an election.
Add Member to Replica-Set
Request:
POST /instances { "instance": { "name": "product-b", ... "topology": { "mongodb": { "type": "member", "replSet": "products", "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", ... } }
Notes:
- If 'join' is true, and there is no existing replica-set for the tenant matching the 'replSet' value, the request will be failed.
- Will have to use 'db.isMaster()' to determine the current primary to execute replica-set commands against (since it can be dynamic due to elections)
- Will use http://docs.mongodb.org/manual/tutorial/expand-replica-set/#configure-and-add-a-member
- Should protect against adding more than 12 members to a replica-set
- Should protect against adding more than 7 voting members to a replica-set
- Should return warning when number of voting members is even and there is no arbiter
Add Another Member to Replica-Set
Request:
POST /instances { "instance": { "name": "product-c", ... "topology": { "mongodb": { "type": "member", "replSet": "products", "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", ... } }
Show Instance
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998
Response:
{ "instance": { ... "topology": { "mongodb": { "type": "member", "replSet": "products" } }, ... } }
Show Topology
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology
Response:
{ "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", ... "mongodb": { "type": "member", "replSet": "products" } }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", ... "mongodb": { "type": "member", "replSet": "products" } }, { "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", "name": "product-c", ... "mongodb": { "type": "member", "replSet": "products" } } ] } }
Add Arbiter
Request:
POST /instances { "instance": { "name": "product-arbiter", ... "topology": { "mongodb": { "type": "arbiter", "replSet": "products", "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "a1b62aaa-7863-4384-8250-59024141c1f8", ... } }
Add a Delayed Member
Request:
POST /instances { "instance": { "name": "product-delayed", ... "topology": { "mongodb": { "type": "member", "replSet": "products", "join": true, "priority": 0, "hidden": true, "slaveDelay": 3600 } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "7d8eb019-931b-4b2a-88d2-4c9f0ca1b29e", ... } }
Notes:
- 'type', 'replSet', 'join', 'priority', 'hidden', and 'slaveDelay' are the only fields supported in topology.mongodb{}. All other configuration values must be set via a configuration-group. After more thought, consider supporting 'hostname' and 'votes' as well.
- Why isn't 'priority', 'hidden' and 'slaveDelay' in a configuration-group you ask? This is explained in "Modifying a Replica-Set" below.
Modifying a Replica-Set
Thus far we've been able to model building a replica-set, adding an arbiter, adding a delayed secondary member, etc. Let's continue with how to modify a replica-set.
Example:
# from http://docs.mongodb.org/manual/tutorial/configure-secondary-only-replica-set-member/#example cfg = rs.conf() cfg.members[0].priority = 2 cfg.members[1].priority = 1 cfg.members[2].priority = 0.5 cfg.members[3].priority = 0 rs.reconfig(cfg)
Executing these priority changes one at a time can have catastrophic results, so it must be done as a transaction (with rs.reconfig() commiting). However, without the ability to address the cluster (i.e. multiple members at once), this becomes impossible. The only backdoor solution would be to guarantee that the MongoDB user(s) presented to the cloud tenant all have the clusterAdmin role, as this would allow them to connect to the primary and execute such transactions themselves via the native client. Obviously however, granting clusterAdmin to every DBaaS user in MongoDB is unacceptable in most deployments.
Option #1: PATCH /instances/:id/topology
Request:
PATCH /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology { "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "mongodb": { "priority": 2 } }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "mongodb": { "priority": 1 } }, { "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", "mongodb": { "priority": 0.5 } } ] } }
Response:
{ "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", ... "mongodb": { "type": "member", "replSet": "products" } }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", ... "mongodb": { "type": "member", "replSet": "products" } }, { "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", "name": "product-c", ... "mongodb": { "type": "member", "replSet": "products" } } ] } }
Notes:
- An HTTP PATCH vs. PUT because the omission of a field or structure should not be an indication to drop/delete it.
- All modified fields in a request will be changed transactionally in a single rs.reconfig().
- It should now be clear why 'priority', 'hidden' and 'slaveDelay' are in topology.mongodb{} vs. a configuration-group: when a configuration-group is changed, an event is immediately triggered to update any attached trove instances. Therefore, if you have a heterogeneous mixture of configuration-groups in a replica-set, there is no way to coordinate a consolidated rs.reconfig().
- Downside: if topology.mongodb{} may have fields returned on a GET that you cannot change in a PATCH/PUT; re-worded, the granularity of what is permissible to change in a PATCH becomes complicated to check and validate.
- TBD on what should be returned in the mongodb{} on a GET /instance/:id and GET /instance/:id/topology. It's a question of whether we should persist anything beyond the 'type' and 'replSet'. If say the 'priority' is stored, you introduce the possibility of drift from the truth, but can easily return it on a GET; if it's not stored, do we prompt MongoDB for the truth on a GET, or is that too computationally expensive?
Option #2: POST /instances/:id/topology/action
Request:
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "mongodb": { "update_members": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "priority": 2 }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "priority": 1 }, { "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", "priority": 0.5 } ] } } }
Notes:
- 'update_members.members[]' elements will only permit 'priority', 'hidden', and 'slaveDelay' (possibly 'votes' and 'hostname' as mentioned earlier).
- Due to the limited field-set, this approach is much more fine-grained than the PATCH approach in Option #1.
Decision: Option #2 is more fine-grained, easier to reason about, and less error-prone. As you'll see in later operations (like Remove a Member), the action vs. PUT/PATCH approach is also more appropriate.
Remove a Member
Removing a member is not the same as deleting one, therefore DELETE /instances/:id is not appropriate.
Option #1: PUT /instances/:id/topology
Request:
PUT /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology { "instance": { "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", } ] } }
Notes:
- By omitting a member{} for id=3a72ee87-cf3e-40f1-a1e1-fe8c7263a782 in a PUT operation, this indicates the member should be removed from the replica-set.
- It's possible that one might want to modify the 'priority', 'hidden', 'votes', etc. fields of the remaining members while dropping a member. So although the example above does not show it, mongodb{} can be included in a member to indicate other changes, *BUT*, since it's a PUT the expectation of what happens to omitted fields in mongodb{} becomes unclear.
Summary: Not very clean, mildly confusing, and very error-prone (nowhere is a "remove" action ever explicitly implied).
Option #2: POST /instances/:id/topology/action
Request:
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "mongodb": { "remove_member": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } }
Notes:
- mongodb{} wrapper isn't necessary, but provides the benefit of schema validation + declaration of intention/understanding.
- The 'remove_member' action is explicit here, vs. implicit as seen in the PUT option.
- 'remove_member' has a strict set of fields that are supported, so there is no question as to what can be provided and what will be honored (as compared to the PUT).
Summary: Fairly clean, with no real drawbacks.
Option #3: POST /instances/:id/topology/remove
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/remove { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" }
Notes:
- Differs from Option #2 in that the action is in the URI vs. the payload.
- One drawback of this approach is that not every action will be supported across all datastores. So for example, a POST /instances/:id/topology/changeoplogsize (http://docs.mongodb.org/manual/tutorial/change-oplog-size/) makes absolutely no sense to any datastore other than MongoDB.
Summary: At first glance is cleaner than Option #2 from a payload-perspective, but the URI discoverability and expansion is awful.
Option #4: POST /instances/:id/action
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/action { "join": false }
Notes:
- Executed against the instance you wish to remove itself from the cluster, so providing the 'id' in the payload is unnecessary.
- Drawback: There are actions that are replica-set-wide (or against a subset of the replica-set), meaning Option #1 or #2 or #3 would have to co-exist with this option anyway.
- Drawback: Increase the number of ways to accomplish the same thing (could unjoin against /instances/:id/action, or against /instances/:id/topology)
Summary: For this very specific example it looks great, but isn't expressive enough for other actions.
Option #5: POST /instances/:id/topology/:id/action
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/dfbbd9ca-b5e1-4028-adb7-f78643e17998/action { "mongodb": { "remove": {} } }
Notes:
- The /instances/:id is an arbitrary member in the replica-set, it doesn't matter which one; the topology/:id is then a member of said replica-set that this action will be applied to.
- Executed against the instance you wish to remove itself from the cluster, so providing the 'id' in the payload is unnecessary.
- Needs More Thought: Could conceivably allow only specific operations here (like remove/unjoin), but not others that could be accomplished in a PATCH against /instances/:id/topology (like 'priority', 'hidden', etc.)
Summary: Fairly clean with no real drawbacks.
Decision: Option #2; it's extremely fine-grained, easy to reason about, and matches the approach taken in Remove A Member.
MongoDB TokuMX
- TokuMUX will require a new datastore-version and *possibly* a new manager class (same reasoning as why Tungsten/Galera will have their own datastore-version for MySQL)
Cassandra
Create Cluster
Request:
POST /instances { "instance": { "name": "product-a", ... "datastore": { "type": "cassandra", "version": "cassandra-2.0.5" }, "topology": { "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": true, "endpoint_snitch": "RackInferringSnitch", "join": false } }, ... } }
Response:
{ "instance": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", ... } }
Notes:
- Unlike in MongoDB, the 'type' field is not required (because all members of the cluster are of the same type)
- 'cluster_name', 'num_tokens', and 'is_seed' are always required, with 'endpoint_snitch' being required if 'join' is false (if 'join' is true, the endpoint_snitch is inherited) and 'auto_bootstrap' required if 'join' is true.
- 'seed_provider' can optionally be provided, but conveniently defaults to 'org.apache.cassandra.locator.SimpleSeedProvider'
- 'join' indicates whether you're joining an existing cluster, or creating a new one. If 'join' is false, and an active cluster by that name for the tenant already exists, the request will be failed. 'join' by default will be false, but was included above for illustrative purposes.
- Likely the Keystone region and the availability-zone inherent to the trove request can be used for the data-center and rack (cassandra-topology.properties), but if it turns out that the naming schemes are incompatible, 'data_center' and 'rack' can be introduced.
- 'is_seed' is used vs. a seed list of ip-addresses because (1) the ip-address is not yet known and (2) when additional seeds are added, each node in the cluster must be notified and updated. More on this later.
Add Node to Cluster
Request:
POST /instances { "instance": { "name": "product-b", ... "topology": { "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": false, "auto_bootstrap": false, "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", ... } }
Notes:
- If 'join' is true, and there is no existing cluster for the tenant matching the 'cluster_name' value, the request will be failed.
- If 'endpoint_snitch' is provided, and the value does not match that of the existing node(s) in the cluster, the request will be failed.
Add Another Node to Cluster
Request:
POST /instances { "instance": { "name": "product-c", ... "topology": { "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": false, "auto_bootstrap": false, "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", ... } }
Show Instance
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998
Response:
{ "instance": { ... "topology": { "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": true } }, ... } }
Show Topology
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology
Response:
{ "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", ... "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": true } }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", ... "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": false } }, { "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", "name": "product-c", ... "cassandra": { "cluster_name": "products", "num_tokens": 256, "is_seed": false } } ] } }
Modifying a Cluster
Example: Drain (http://www.datastax.com/documentation/cassandra/2.0/cassandra/tools/toolsDrain.html)
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "cassandra": { "drain_node": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } }
Remove a Member
Request:
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "cassandra": { "remove_node": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } }
Couchbase
Create Cluster
Create Initial Cluster
Request:
POST /instances { "instance": { "name": "product-a", ... "datastore": { "type": "couchbase", "version": "couchbase-2.2" }, "topology": { "couchbase": { "cluster_name": "products", "join": false } }, ... } }
Response:
{ "instance": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", ... } }
Notes:
- Couchbase does not support a "cluster name", but does require a xdcr-cluster-name when using XDCR.
- Tip: Summary of operations at http://docs.couchbase.com/couchbase-manual-2.2/#xdcr-replicate-options
Add Node to Cluster
Request:
POST /instances { "instance": { "name": "product-b", ... "topology": { "couchbase": { "cluster_name": "products", "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", ... } }
Notes:
- If 'join' is true, and there is no existing cluster for the tenant matching the 'cluster_name' value, the request will be failed.
- In Couchbase, a new node can join the cluster by referencing any existing member, which is discoverable by trove via the cluster_name (unique per tenant)
Add Another Node to Cluster
Request/Response omitted due to a lack of any special considerations required
Show Instance
Request/Response omitted due to a lack of any special considerations required
Show Topology
Request/Response omitted due to a lack of any special considerations required
Modifying a Cluster
POST /instances/:id/topology/action
Example: Create Bucket
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "couchbase": { "create_bucket": { "bucket": "test_bucket", "bucket_type": "couchbase", "bucket_port": 11222, "bucket_ramsize": 200, "bucket_replica": 1 } } }
Remove a Member
POST /instances/:id/topology/action
Request:
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "couchbase": { "rebalance": { "server_remove": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } }
Notes:
- Note how removing a node is done in the context of a rebalance for Couchbase. This is an example of how homogenizing actions across trove, like "remove node", might not be appropriate.
Redis
Create Master
Request:
POST /instances { "instance": { "name": "product-a", ... "datastore": { "type": "redis", "version": "redis-2.8.6" }, "topology": { "redis": {} }, ... } }
Response:
{ "instance": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", ... } }
Notes:
- 'join' field is not required for redis. also notice the lack of a 'cluster_name' of sorts (see 'Add Slave' for reasoning)
Add Slave
Request:
POST /instances { "instance": { "name": "product-b", ... "datastore": { "type": "redis", "version": "redis-2.8.6" }, "topology": { "redis": { "slave_of": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", ... } }
Notes:
- redis supports daisy-chaining slaves, therefore the 'slave_of' value needs to be a specific trove instance uuid vs. a manufactured 'cluster_name' of sorts.
- whether it is a master or slave can be inferred from the presence (or lack thereof) of 'slave_of'.
Show Instance
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998
Response:
{ "instance": { ... "topology": { "redis": {} }, ... } }
Show Topology
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology
Response:
{ "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", ... "redis": {} }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", ... "redis": { "slave_of": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } ] } }
Promote/Disconnect Slave
POST /instances/:id/topology/action
Request:
POST /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology/action { "redis": { "slaveof_no_one": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } }
Redis Cluster
Create Cluster
Request:
POST /instances { "instance": { "name": "product-a", ... "datastore": { "type": "redis", "version": "redis-3.0.0-beta1" }, "topology": { "redis_cluster": { "cluster_name": "products", "cluster_timeout": 5000, "join": false } }, ... } }
Response:
{ "instance": { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", ... } }
Notes:
- Will require datastore-version to topology.<namespace> validation (e.g. if redis 3+, permit topology.redis_cluster{} and topology.redis{}, but if redis 2.x only permit topology.redis{})
- 'cluster_name' and 'cluster_timeout' are unique to topology.redis_cluster{} (not supported in topology.redis{})
- As seen in other datastore examples, uses 'join' and 'cluster_name'
Add Another Master to Cluster
Request:
POST /instances { "instance": { "name": "product-b", ... "datastore": { "type": "redis", "version": "redis-3.0.0-beta1" }, "topology": { "redis_cluster": { "cluster_name": "products", "join": true } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", ... } }
Add Slave to Cluster
Request:
POST /instances { "instance": { "name": "product-c", ... "datastore": { "type": "redis", "version": "redis-3.0.0-beta1" }, "topology": { "redis_cluster": { "cluster_name": "products", "slave_of": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } }, ... } }
Response:
{ "instance": { "status": "BUILD", "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", ... } }
Notes:
- Despite redis clusters supporting the ability to add a slave without designating the master, we will require it to avoid unoptimal geographical relationships.
Show Instance
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998
Response:
{ "instance": { ... "topology": { "redis_cluster": { "cluster_name": "products", "cluster_timeout": 5000 } }, ... } }
Show Topology
Request:
GET /instances/dfbbd9ca-b5e1-4028-adb7-f78643e17998/topology
Response:
{ "topology": { "members": [ { "id": "dfbbd9ca-b5e1-4028-adb7-f78643e17998", "name": "product-a", ... "redis_cluster": { "cluster_name": "products", "cluster_timeout": 5000 } }, { "id": "061aaf4c-3a57-411e-9df9-2d0f813db859", "name": "product-b", ... "redis_cluster": { "cluster_name": "products", "cluster_timeout": 5000 } }, { "id": "3a72ee87-cf3e-40f1-a1e1-fe8c7263a782", "name": "product-c", ... "redis_cluster": { "cluster_name": "products", "cluster_timeout": 5000, "slave_of": "dfbbd9ca-b5e1-4028-adb7-f78643e17998" } } ] } }
Promote/Disconnect Slave
Work in Progress
Feedback
Masters/Slaves
- drewford: Can one make a master/slave out of an existing instance?
- drewford: Type and version should follow the same schema as listing datastore types
Arbiters
- drewford: Could/should arbiters be managed by the system automatically? Is it worth putting that responsibility on the API user, or can it be automated by the system for an easier, less technical experience? From reading the article noted in the link, it seems like it's a good enough idea to automate it to reduce the need for users to research and decide for themselves.
Delayed Members
- drewford: Could/should delayed members be managed by the system automatically? Can it just be fully automated for a better, less technical experience?
Topology
- drewford: Members of the topology in the mongo example all have the same "mongodb" object. Unless this could also be "mysql" or "redis" in the same topology, it might make sense to move up "type" and "replSet" attributes as siblings to "name".
- drewford: The "type" attribute also seems redundant to the name of the array. Are there attributes other than "members" on a "topology"? If not, can "topology" be the array with each object having a "type"?
Cassandra Create Cluster
- drewford: can "num_tokens", "endpoint_snitch" or "auto_bootstrap" be automated in any way so they receive a valid default value even if the user does not specify one? Looking for ways in which we can take the decision workload off the user, and put it on the system in a healthy way.
Promote/Disconnect Redis Slave
- drewford: Seems like the promotion/disconnection json should be structured the same across all datastores if possible - with the backend handling any special manipulation to get the job done. Standardize in favor of the user.
Redis Cluster - Create Cluster
- drewford: "redis_cluster" here is labelled differently than the same object in Cassandra and Couchbase - they don't use "_cluster".