manageiq / inventory_refresh Goto Github PK

When there were multiple strategies (topological and recursive) it made sense to have scanning separate but now it can live under topological_sort

The 90% case is that these are a single column (e.g. :ems_ref or :uid_ems) so allow the user to pass in a single value and wrap it in an array when building the collection.

Removing ManageIQ's persister and builder

• #26 (move them to spec/helpers)
  • #27
    • #29

--- a/lib/inventory_refresh/save_collection/saver/base.rb
+++ b/lib/inventory_refresh/save_collection/saver/base.rb
@@ -163,7 +163,7 @@ module InventoryRefresh::SaveCollection
               # Record was found in the DB but not sent for saving, that means it doesn't exist anymore and we should
               # delete it from the DB.
               delete_record!(record) if inventory_collection.delete_allowed?
-            else
+            elsif !inventory_collection.create_only?
               # Record was found in the DB and sent for saving, we will be updating the DB.
               update_record!(record, hash, inventory_object) if assert_referential_integrity(hash)

The create_only? option is honored by concurrent_safe_batch and batch save strategies but not by base.

• secondary_ref is an odd name choice...perhaps extra_indexes, alternate_indexes, lookups?
• When you use a secondary_ref, the current syntax is

# collection def
:secondary_refs = {:by_name => %i(name)}

# in parser
collection.lazy_find({:name => "foo"}, {:ref => :by_name})

A more ergnomic syntax would be

collection.lazy_find(:by_name => {:name => "foo"})

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Ignored or Blocked

These are blocked by an existing closed PR and will not be recreated unless you click a checkbox below.

• Update dependency activerecord to v6 [SECURITY]
• Update dependency activerecord to "~>6.1.0"
• Update dependency activerecord to v7

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

There are use cases where we want to assign default values to new records but do not overwrite these values on updated records e.g. MiqWidget Seeding on the enabled and visibility columns.

We're hitting an issue where we have to copy value from Flavor to Vm, but divide it by 2^20 prior saving. Something like:

persister.flavors.build(
  ... 
  :memory => flavor.memory_in_mb.megabytes # store in bytes
)

persister.hardwares.build(
  ...
  :memory_mb => persister.flavors.find(flavor_ref).memory / 1.megabytes  # convert to megabytes
)

Problem is that we currently have to use .find() which may potentially result in ordering issues. Would it be possible to let .lazy_find() accept a block where we could preprocess the value? Something like this would be fantastic:

persister.flavors.lazy_find(flavor_ref) { |flavor| flavor.memory / 1.megabyte }

/cc @agrare

https://github.com/ManageIQ/inventory_refresh/blob/master/lib/inventory_refresh/target.rb#L57