Migration scripts for OAI-PMH

Populate complete_updated_date column in instance table

0. Disable trigger

Trigger should be disabled to not override historical data to be populated in `complete_updated_date` column

ALTER TABLE {tenant}_mod_inventory_storage.instance DISABLE TRIGGER updatecompleteupdateddate_instance;

1. Prepare step

On this step uuid_range table is created in public schema to split UUIDs in different partitions (16 partitions in total) and track their status later (completed column).

create table public.uuid_range
    id             serial primary key,
    partition      uuid,
    subrange_start uuid,
    subrange_end   uuid,
    completed      boolean default false

        partition_start UUID;
        subrange_start  UUID;
        subrange_end    UUID;
        FOR i IN 0..15
                partition_start := (rpad(to_hex(i), 8, '0') || '-0000-0000-0000-000000000000')::UUID;
                subrange_start := partition_start;

                FOR j IN 1..4096
                        IF i < 15 OR (i = 15 AND j < 4096) THEN
                            if (j < 4096) then
                                subrange_end := (to_hex(i)::text || rpad(lpad(to_hex(j), 3, '0'), 7, '0') ||
                                subrange_end := (rpad(to_hex(i + 1), 8, '0') || '-0000-0000-0000-000000000000')::UUID;
                            end if;
                            subrange_end := 'ffffffff-ffff-ffff-ffff-ffffffffffff'::UUID; -- upper bound for last subrange in last partition
                        END IF;

                        INSERT INTO public.uuid_range (partition, subrange_start, subrange_end)
                        VALUES (partition_start, subrange_start, subrange_end);

                        subrange_start := subrange_end;
                    END LOOP;
            END LOOP;
LANGUAGE plpgsql; 

2. Update

Update step should be executed for each partition created from step #1. Update script can be executed concurrently (tests results can be found in PERF-725 - Getting issue details... STATUS to understand how many partitions can be processed at the same time on specific env without users impact)

Params to be updated 
tenant - tenant name

partition - partition value from public.uuid_range table

set search_path = "{tenant}_mod_inventory_storage", "public";

        partition_to_process uuid := '{partition}'::uuid;
        range_rec           public.uuid_range%rowtype;
        for range_rec in select ur.*
                         from public.uuid_range ur
                         where ur.partition = partition_to_process
                           and ur.completed = false
                --             start transaction;
                UPDATE {tenant}_mod_inventory_storage.instance i
                SET complete_updated_date = (jsonb -> 'metadata' ->> 'updatedDate')::timestamp with time zone
                where i.id between range_rec.subrange_start and range_rec.subrange_end;

                update public.uuid_range
                set completed = true
                where id = range_rec.id;
                raise notice 'subrange id: % completed', range_rec.id;
            end loop;

3. Monitor

Next script can be used to monitor how many sub-partitions in partitions are left to be processed

select partition, count(partition)
from public.uuid_range
where completed = false
group by partition
order by partition;   

4. Post-migration

As initially trigger was disabled, it should be enabled again, additionally index should be created for better performance

ALTER TABLE {tenant}_mod_inventory_storage.instance ENABLE TRIGGER updatecompleteupdateddate_instance;

CREATE INDEX IF NOT EXISTS idx_instance_complete_updated_date
ON {tenant}_mod_inventory_storage.instance(complete_updated_date);