生产环境有条慢sql需优化，耗时609s,如下：

	sum( fsl_unrelease_pack_ob_lines.qty ) AS qty_sum,
	sum( fsl_unrelease_pack_ob_lines.box_qty ) AS box_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.release_qty ) AS release_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.box_qty - fsl_unrelease_pack_ob_lines.release_qty ) AS unrelease_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.box_qty - fsl_unrelease_pack_ob_lines.release_qty ) AS add_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.weight ) AS weight_sum,
	sum( fsl_unrelease_pack_ob_lines.volume ) AS volume_sum 
FROM
	fsl_order_base_line AS fsl_unrelease_pack_ob_lines
	LEFT JOIN fsl_order_base AS ob ON fsl_unrelease_pack_ob_lines.order_base = ob.id 
WHERE
	ob.model1 = 'pack' 
	AND ob.project_code = 'DD' 
	AND fsl_unrelease_pack_ob_lines.domain_name IN ( 'FSL' ) 
	AND fsl_unrelease_pack_ob_lines.origin IN (
	SELECT
		location 
	FROM
		fsl_zone_part t1
		JOIN fsl_location t2 ON t1.zone1 = t2.id 
	WHERE
	t2.NAME = 'xx')

》》 查看执行计划如下：

》》表数据量如下：

》》考虑优化：改写sql+强制索引
这里是考虑小表驱动大表
改写后的sql如下,需强制索引才走对应的执行计划

	sum( fsl_unrelease_pack_ob_lines.qty ) AS qty_sum,
	sum( fsl_unrelease_pack_ob_lines.box_qty ) AS box_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.release_qty ) AS release_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.box_qty - fsl_unrelease_pack_ob_lines.release_qty ) AS unrelease_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.box_qty - fsl_unrelease_pack_ob_lines.release_qty ) AS add_qty_sum,
	sum( fsl_unrelease_pack_ob_lines.weight ) AS weight_sum,
	sum( fsl_unrelease_pack_ob_lines.volume ) AS volume_sum 
FROM
   fsl_zone_part t1 JOIN fsl_location t2 ON t1.zone1 = t2.id
	join 
	fsl_order_base_line AS fsl_unrelease_pack_ob_lines force index(index_obl_origin) on fsl_unrelease_pack_ob_lines.origin=t1.location
	LEFT JOIN fsl_order_base AS ob ON fsl_unrelease_pack_ob_lines.order_base = ob.id 
WHERE  t2.NAME = '华南'
	 AND ob.model1 = 'pack' 	
				AND  ob.project_code = 'DD' 
	AND fsl_unrelease_pack_ob_lines.domain_name IN ( 'FSL' ) 

执行计划：

但耗时仍需400s，请问下有进一步优化空间？