diff --git a/rados/operation_test.go b/rados/operation_test.go
new file mode 100644
index 0000000..67c6cc8
--- /dev/null
+++ b/rados/operation_test.go
@@ -0,0 +1,119 @@
+package rados
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestOperationError(t *testing.T) {
+	oe := OperationError{
+		kind:    readOp,
+		OpError: fmt.Errorf("bad mojo %v", 77),
+		StepErrors: map[int]error{
+			1: fmt.Errorf("limit exceeded"),
+			3: fmt.Errorf("weirdness"),
+		},
+	}
+	assert.Error(t, oe)
+	estr := oe.Error()
+	assert.Contains(t, estr, "read operation error")
+	assert.Contains(t, estr, "op=bad mojo 77")
+	assert.Contains(t, estr, "Step#1=limit exceeded")
+	assert.Contains(t, estr, "Step#3=weirdness")
+
+	oe = OperationError{
+		kind: writeOp,
+		StepErrors: map[int]error{
+			0: fmt.Errorf("unlucky"),
+		},
+	}
+	assert.Error(t, oe)
+	estr = oe.Error()
+	assert.Contains(t, estr, "write operation error")
+	assert.NotContains(t, estr, "op=")
+	assert.Contains(t, estr, "Step#0=unlucky")
+}
+
+type fooStep struct {
+	updateCount int
+	freeCount   int
+	failMe      error
+}
+
+func (fs *fooStep) update() error {
+	fs.updateCount++
+	return fs.failMe
+}
+
+func (fs *fooStep) free() {
+	fs.freeCount++
+}
+
+func TestOpStepFinalizer(t *testing.T) {
+	// this confirms that given a valid op step the finalizer helper
+	// function calls the free method of the interface
+	fs := &fooStep{}
+	opStepFinalizer(fs)
+	opStepFinalizer(fs)
+	assert.Equal(t, 2, fs.freeCount)
+
+	// this should not panic
+	opStepFinalizer(nil)
+}
+
+func TestOperationType(t *testing.T) {
+	t.Run("stepErrors", func(t *testing.T) {
+		o := &operation{}
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{failMe: fmt.Errorf("yow")})
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{failMe: fmt.Errorf("ouch")})
+
+		err := o.update(readOp, 0)
+		if assert.Error(t, err) {
+			oe := err.(OperationError)
+			assert.NoError(t, oe.OpError)
+			assert.Len(t, oe.StepErrors, 2)
+			assert.Contains(t, oe.StepErrors, 1)
+			assert.Contains(t, oe.StepErrors, 4)
+		}
+	})
+
+	t.Run("opError", func(t *testing.T) {
+		o := &operation{}
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{})
+
+		err := o.update(readOp, 5)
+		if assert.Error(t, err) {
+			oe := err.(OperationError)
+			assert.Error(t, oe.OpError)
+			assert.Len(t, oe.StepErrors, 0)
+		}
+	})
+
+	t.Run("noErrors", func(t *testing.T) {
+		o := &operation{}
+		o.steps = append(o.steps, &fooStep{})
+		o.steps = append(o.steps, &fooStep{})
+		err := o.update(readOp, 0)
+		assert.NoError(t, err)
+		x := 0
+		for i := range o.steps {
+			x += o.steps[i].(*fooStep).updateCount
+		}
+		assert.Equal(t, 2, x)
+
+		o.free()
+		x = 0
+		for i := range o.steps {
+			x += o.steps[i].(*fooStep).updateCount
+		}
+		assert.Equal(t, 2, x)
+	})
+}