Commit Graph

25 Commits

Author SHA1 Message Date
Chris S. Kim 0a6d683c84
Update Intentions.List with partitions (#11299) 2021-10-13 10:47:12 -04:00
R.B. Boyer d730298f59
state: partition the nodes.uuid and nodes.meta indexes as well (#10882) 2021-08-19 16:17:59 -05:00
R.B. Boyer e50e13d2ab
state: partition nodes and coordinates in the state store (#10859)
Additionally:

- partitioned the catalog indexes appropriately for partitioning
- removed a stray reference to a non-existent index named "node.checks"
2021-08-17 13:29:39 -05:00
R.B. Boyer 254557a1f6
sync changes to oss files made in enterprise (#10670) 2021-07-22 13:58:08 -05:00
Daniel Nephin d785c86db1 state: convert checks.service index to new pattern 2021-03-29 16:38:53 -04:00
Daniel Nephin f859ba6d4b state: convert checks.status indexer
As part of this change the indexer will now be case insensitive by using
the lower case value. This should be safe because previously we always
had lower case strings.

This change was made out of convenience. All the other indexers use
lowercase, so we can re-use the indexFromQuery function by using
lowercase here as well.
2021-03-29 16:38:50 -04:00
Daniel Nephin 372d274b34 state: convert services.kind to functional indexer pattern 2021-03-29 15:42:30 -04:00
Daniel Nephin 9f9eadd569 state: convert services table service and connect indexer
To the new functional indexer pattern
2021-03-29 15:42:24 -04:00
Daniel Nephin ec04df66bd state: remove duplication of Query indexer 2021-03-29 14:35:11 -04:00
Daniel Nephin c6a1ca701d state: reduce duplication in catalog table schema 2021-03-29 14:21:23 -04:00
Daniel Nephin 25b791ba47 state: add tests for checks.ID indexer 2021-03-22 18:06:43 -04:00
Daniel Nephin 49938bc472 state: convert checks.ID index to new pattern 2021-03-22 18:06:08 -04:00
Daniel Nephin bae69b2352 state: fix prefix index with the new pattern
Prefix queries are generally being used to match part of a partial
index. We can support these indexes by using a function that accept
different types for each subset of the index.

What I found interesting is that in the generic StringFieldIndexer the
implementation for PrefixFromArgs would remove the trailing null, but
at least in these 2 cases we actually want a null terminated string.
We simply want fewer components in the string.
2021-03-19 14:12:17 -04:00
Daniel Nephin ec50454fb3 state: move services.ID to new pattern 2021-03-19 14:11:59 -04:00
Daniel Nephin 66632538d8 state: use constants and remove wrapping
for GatewayServices table
2021-03-18 12:08:59 -04:00
Daniel Nephin d77bdd26c5 state: Move UpstreamDownstream to state package 2021-03-18 12:08:59 -04:00
Daniel Nephin 8a1a11814d state: use constants for mesh-topology table operations 2021-03-18 12:08:03 -04:00
Daniel Nephin 0b3930272d state: convert services.node and checks.node indexes
Using NodeIdentity to share the indexes with both.
2021-03-16 13:00:31 -04:00
Daniel Nephin 1184ceff9e state: convert nodes.ID to new functional pattern
In preparation for adding other identifiers to the index.
2021-03-05 12:30:40 -05:00
Daniel Nephin 7e4d693aaa state: support for functional indexers
These new functional indexers provide a few advantages:

1. enterprise differences can be isolated to a single function (the
   indexer function), making code easier to change
2. as a consequence of (1) we no longer need to wrap all the calls to
   Txn operations, making code easier to read.
3. by removing reflection we should increase the performance of all
   operations.

One important change is in making all the function signatures the same.

https://blog.golang.org/errors-are-values

An extra boolean return value for SingleIndexer.FromObject is superfluous.
The error value can indicate when the index value could not be created.
By removing this extra return value we can use the same signature for both
indexer functions.

This has the nice properly of a function being usable for both indexing operations.
2021-02-19 17:14:46 -05:00
Daniel Nephin 88a9bd6d3c state: remove duplicate index on the checks table
By using a new pattern for more specific indexes. This allows us to use
the same index for both service checks and node checks. It removes the
abstraction around memdb.Txn operations, and isolates all of the
enterprise differences in a single place (the indexer).
2021-02-19 17:14:46 -05:00
Daniel Nephin b747b27afd state: remove the need for registerSchema
registerSchema creates some indirection which is not necessary in this
case. newDBSchema can call each of the tables.

Enterprise tables can be added from the existing withEnterpriseSchema
shim.
2021-02-05 12:19:56 -05:00
Daniel Nephin afdbf2a8ef state: rename table name constants to new pattern
Using Apps Hungarian Notation for these constants makes the memdb queries more readable.
2021-02-05 12:12:18 -05:00
Daniel Nephin 825b8ade39 state: use indexID
this change was already made to enterprise, so backporting it.
2021-01-28 20:30:08 -05:00
Daniel Nephin aa21c1ea04 state: reduce interface for Enterprise schema
Using withEnterpriseSchema() we can apply any enterprise schema changes
with a single shim, removing the need to duplicate all of the table
definitions.

Also move all the catalog schemas to a new file to shrink catalog.go a bit.
2021-01-15 18:49:55 -05:00