Over the weekend, the N4Free server lost its iSCSI and of course all the VMs crashed and lost data.
In the log I see this at the time we lost iSCSI:
Code: Select all
Apr 26 03:11:20 fractal istgt[1936]: ABORT_TASK
Apr 26 03:11:55 fractal istgt[1936]: ABORT_TASK
Apr 26 03:12:06 fractal istgt[1936]: Login from iqn.1998-01.com.vmware:esx14-749bb95e (172.30.1.14) on iqn.2007-09.uk.co.bdl.fractal:vm1targ LU1 (172.30.1.16:3260,1), ISID=23d000002, TSIH=39, CID=0, HeaderDigest=off, DataDigest=off
Apr 26 03:12:06 nas4free istgt[1936]: Login from iqn.1998-01.com.vmware:esx14-749bb95e (172.30.1.14) on iqn.2007-09.uk.co.bdl.fractal:vm1targ LU1 (172.30.1.16:3260,1), ISID=23d000002, TSIH=39, CID=0, HeaderDigest=off, DataDigest=off
Apr 26 03:13:04 fractal istgt[1936]: ABORT_TASK
Apr 26 03:14:25 fractal istgt[1936]: istgt_iscsi.c: 766:istgt_iscsi_read_pdu: ***ERROR*** readv() failed (-1,errno=35,iqn.1998-01.com.vmware:esx8-014bc0db,time=41)
Apr 26 03:14:25 fractal istgt[1936]: istgt_iscsi.c:5685:worker: ***ERROR*** conn->state = 1
Apr 26 03:14:25 fractal istgt[1936]: istgt_iscsi.c:5702:worker: ***ERROR*** iscsi_read_pdu() failed
Apr 26 03:14:56 nas4free istgt[1936]: Login from iqn.1998-01.com.vmware:esx8-014bc0db (193.195.25.8) on iqn.2007-09.uk.co.bdl.fractal:vm1targ LU1 (193.195.25.89:3260,1), ISID=23d000001, TSIH=40, CID=0, HeaderDigest=off, DataDigest=off
Apr 26 03:14:56 fractal istgt[1936]: Login from iqn.1998-01.com.vmware:esx8-014bc0db (193.195.25.8) on iqn.2007-09.uk.co.bdl.fractal:vm1targ LU1 (193.195.25.89:3260,1), ISID=23d000001, TSIH=40, CID=0, HeaderDigest=off, DataDigest=off
I found the following related threads, where the spec of the server is called into question. However this box is fairly beefy:
Quad Core Intel I5
32GB RAM (tuned for 24GB using the ZFS extension)
SSD read cache (120GB)
6 x WD Red 3TB in mirrored pairs (all working in ZFS status and no SMART errors)
NAS4Free vers "9.2.0.1 - Shigawire (revision 925)"
other threads
viewtopic.php?f=80&t=4852
viewtopic.php?f=33&t=5304
