diff --git a/lib/ftl/ftl_restore.c b/lib/ftl/ftl_restore.c
index e269a0537..a5f241ebf 100644
--- a/lib/ftl/ftl_restore.c
+++ b/lib/ftl/ftl_restore.c
@@ -180,7 +180,6 @@ ftl_restore_check_seq(const struct ftl_restore *restore)
 
 	for (i = 0; i < ftl_dev_num_bands(dev); ++i) {
 		rband = &restore->bands[i];
-
 		if (rband->md_status != FTL_MD_SUCCESS) {
 			continue;
 		}
@@ -533,7 +532,14 @@ ftl_restore_pad_band(struct ftl_restore_band *rband)
 		ppa = band->chunk_buf[i].start_ppa;
 		ppa.lbk = info.wp;
 
-		buffer = spdk_dma_zmalloc(FTL_BLOCK_SIZE * dev->xfer_size, sizeof(uint32_t), NULL);
+		/*
+		 * We need 4k alignment for lightnvm writes; otherwise, due to a bug in QEMU,
+		 * scatter gather lists to underlying block device become broken as the number of
+		 * incoming offsets (dev->xfer_size) would be smaller than the calculated size of
+		 * sgl (dev->xfer_size+1), which eventually results in some part of the write
+		 * hitting offset 0 of the drive, overwriting head_md of band 0.
+		 */
+		buffer = spdk_dma_zmalloc(FTL_BLOCK_SIZE * dev->xfer_size, FTL_BLOCK_SIZE, NULL);
 		if (spdk_unlikely(!buffer)) {
 			rc = -ENOMEM;
 			goto error;