ResumeAll doesn't resume 'aborting' transactions

[John A Meinel]

One bit of digging from the production controller is that they have a few (~500) transactions that are in state 3 == aborting, some of which are more than a year old.

Looking into ResumeAlL():
func (r *Runner) ResumeAll() (err error) {
 debugf("Resuming all unfinished transactions")
 iter := r.tc.Find(bson.D{{"s", bson.D{{"$in", []state{tpreparing, tprepared, tapplying}}}}}).Iter()

That means it looks for 1, 2, 4, but not state 3.

One of these *will* get applied if another transaction touches the document that has an aborting transaction, so they probably aren't serious, but it does mean that we have a growing number of transactions that won't ever be progressed to Aborted, which means they won't get pruned.

It may be as simple as just adding taborting to the above list.

[Canonical Juju QA Bot (juju-qa-bot)]

This bug has not been updated in 2 years, so we're marking it Low importance. If you believe this is incorrect, please update the importance.