I have noticed recently, that new nms:recipient records get assigned @id value over 300 million. Currently, the larges @id for this entity is 383,349,601 and it rising rapidly.
Obviously, we do not have so many customers, the real number is much smaller. I have done a quick investigation, and it seems that nms:recipient entity actually shares identity numbers with xtk:workflow and few others entities. When any of these records are created, the id is picked up from the same autoincrement value.
Now, problem is, we use a lot of sub-workflow or some temporary workflows, that gets spawn, executed and then deleted. We will have a couple millions of these created per day, so this auto-increment value will eventually overfill. What should we do? Is it possible to assign some custom @id to a workflow when spawning it, instead of picking an autoincrement value? Or Adobe will, actually, check if record with such id already exists before assigning it?
If you are using autopk="true" or pkSequence="xtkNewId" then yes both tables are using same sequence.
To resolve issue of sequence filling up fast you can do:
1) Define custom sequence and use the same in your table. or
2) Find gaps in your sequence and point your sequence to start from that point.
If you are using sequence available in campaign, all the records that are generated across all the schemas use the same sequence and that might be the reason why you ended using using the sequence. we generally tend up using all the ID's for logging and running workflows temporarily which use the sequence. you can have custom sequence set up for the schemas that consume lot of ID's and there is some material available like the link below.