Configurable initial and max retained batch sizes

[lahsivjar]

Closes: #2780

[cockroach-teamcity]

This change is 

[lahsivjar]

@jbowens Would appreciate a quick review whenever you have time.

[lahsivjar]

@jbowens Friendly ping, any feedback on the PR is appreciated.

[inge4pres]

This is very cool and I also have a similar use case 👍🏼
Plus, there is no breaking change to the public API so it would be nice to have this landed 🙏🏼

[jbowens]

Sorry for the delay here! 😅

Reviewable status: 0 of 3 files reviewed, 1 unresolved discussion (waiting on @lahsivjar)

---

batch.go line 275 at r4 (raw file):

	formatKey      base.FormatKey
	abbreviatedKey AbbreviatedKey
	opts           *batchOptions

I think this should not be a pointer—otherwise, we'll suffer an additional heap allocation for every new batch initialization

[RaduBerinde]

Reviewable status: 0 of 3 files reviewed, 2 unresolved discussions (waiting on @jbowens and @lahsivjar)

---

batch.go line 275 at r4 (raw file):

Previously, jbowens (Jackson Owens) wrote…

I think this should not be a pointer—otherwise, we'll suffer an additional heap allocation for every new batch initialization

+1. ensureDefaults should just modify in-place and return nothing.

---

batch.go line 2421 at r4 (raw file):

func WithInitialSizeBytes(s int) BatchOption {
	return func(opts *batchOptions) {
		if s > 0 {

[nit] The special handling of s <= 0 is odd here, it shouldn't be valid to use such a value.

[lahsivjar]

@jbowens @RaduBerinde Thanks both for your reviews. I have addressed the comments now.

[RaduBerinde]

Thanks! Can you squash the commits?

Reviewable status: 0 of 3 files reviewed, 3 unresolved discussions (waiting on @jbowens and @lahsivjar)

---

batch_test.go line 1744 at r5 (raw file):

		},
		{
			name: "with_invalid_values",

I think we need to remove this test.

[lahsivjar]

Thanks! Can you squash the commits?

@RaduBerinde Done! Also removed the bad test.

[RaduBerinde]

Thank you!