FC data overflow problem

[slushieken]

I have been trying to use FC to my xenserver to have ZFS as a SAN. This works well most of the time, but if freenas goes under load, like in the case of copying a VM from one LUN to another, the following error occurs and takes down the link.

Jan 14 09:51:02 vault ctlfeasync: WWPN 0x2100001b321384ed port 0x0000ef path 0 target 0 arrived
Jan 14 10:08:18 vault ATP: [0x0] origdlen 0 bytes_xfrd 0 lun 0 nphdl 0x0000 s_id 0x000000 d_id 0x000000 oxid 0x0000 state Free
Jan 14 10:08:18 vault (noperiph:isp0:0:0:0): ATP: [0x0] origdlen 0 bytes_xfrd 0 lun 0 nphdl 0x0000 s_id 0x000000 d_id 0x000000 oxid 0x0000 state Free
...
Jan 14 10:08:18 vault (noperiph:isp0:0:0:0): ATP: [0x0] origdlen 0 bytes_xfrd 0 lun 0 nphdl 0x0000 s_id 0x000000 d_id 0x000000 oxid 0x0000 state Free
Jan 14 10:08:18 vault isp0: isp_target_start_ctio: [0x1231f0] data overflow by 4096 bytes

I checked the source code to see if anything made sense:

1712 /*
1713 * Check for overflow
1714 */
1715 tmp = atp->bytes_xfered + atp->bytes_in_transit + xfrlen;
1716 if (tmp > atp->orig_datalen) {
1717 isp_prt(isp, ISP_LOGERR, "%s: [0x%x] data overflow by %u bytes", __func__, cso->tag_id, tmp - atp->orig_datalen);
1718 ccb->ccb_h.status = CAM_DATA_RUN_ERR;
1719 xpt_done(ccb);
1720 continue;
1721 }

Could anyone provide any guidance or hints? I don't know where to go with this at all...

It could be triggered by mismatched block sizes. I have set everything to default on those.

 

[dlavigne]

Please create a bug report at bugs.freenas.org and post the issue number here.

 

[slushieken]

https://bugs.freenas.org/issues/7505 submitted... Thank you