External storage on rpi2 stops working every once in a week or so

I have a raspbian on rpi2 with an external hard disk connected to it through USB. The hard disk is a 1TB seagate with its own power supply.

The hard disk is mounted locally and shared via nfs.

The problem is every once in while (once a week or so), the hard disk seems to stop responding. When it happens I see on my main computer (kubuntu 18.04) that the nfs share does not reply anymore. when I ssh to the rpi machine I see that also there I can't access the hard disk.

I have to power off/on the hard disk and reboot the rpi machine to get it all working again.

My question is, how to go about and fix this problem? what is the method for discovering what's going on?

UPDATE:

so now that it happened again, I run dmesg, and here is what I found (the hard disk is sda):

[   19.565607] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null)
[42530.347552] usb 1-1.2: reset high-speed USB device number 4 using dwc_otg
[51845.323983] sd 0:0:0:0: [sda] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
[51845.324002] sd 0:0:0:0: [sda] tag#0 Sense Key : 0x2 [current] 
[51845.324012] sd 0:0:0:0: [sda] tag#0 ASC=0x4 ASCQ=0x2 
[51845.324026] sd 0:0:0:0: [sda] tag#0 CDB: opcode=0x28 28 00 38 c0 09 08 00 00 08 00
[51845.324037] print_req_error: I/O error, dev sda, sector 952109320
[71399.355286] usb 1-1.2: reset high-speed USB device number 4 using dwc_otg
[77157.409748] usb 1-1.2: reset high-speed USB device number 4 using dwc_otg
[77289.129226] INFO: task usb-storage:69 blocked for more than 120 seconds.
[77289.129242]       Tainted: G         C      4.14.62-v7+ #1134
[77289.129248] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[77289.129257] usb-storage     D    0    69      2 0x00000000
[77289.129312] [<8079d258>] (__schedule) from [<8079d8c0>] (schedule+0x50/0xa8)
[77289.129337] [<8079d8c0>] (schedule) from [<807a14a4>] (schedule_timeout+0x1d0/0x3f4)
[77289.129359] [<807a14a4>] (schedule_timeout) from [<8079e544>] (wait_for_common+0xc0/0x184)
[77289.129378] [<8079e544>] (wait_for_common) from [<8079e628>] (wait_for_completion+0x20/0x24)
[77289.129398] [<8079e628>] (wait_for_completion) from [<805b5724>] (usb_sg_wait+0x174/0x188)
[77289.129424] [<805b5724>] (usb_sg_wait) from [<805ed43c>] (usb_stor_bulk_transfer_sglist.part.2+0x88/0xec)
[77289.129448] [<805ed43c>] (usb_stor_bulk_transfer_sglist.part.2) from [<805ed4fc>] (usb_stor_bulk_srb+0x5c/0x64)
[77289.129470] [<805ed4fc>] (usb_stor_bulk_srb) from [<805ed628>] (usb_stor_Bulk_transport+0x124/0x370)
[77289.129492] [<805ed628>] (usb_stor_Bulk_transport) from [<805edeec>] (usb_stor_invoke_transport+0x30/0x4bc)
[77289.129514] [<805edeec>] (usb_stor_invoke_transport) from [<805ecbf0>] (usb_stor_transparent_scsi_command+0x18/0x1c)
[77289.129534] [<805ecbf0>] (usb_stor_transparent_scsi_command) from [<805ef3bc>] (usb_stor_control_thread+0x198/0x2a0)
[77289.129554] [<805ef3bc>] (usb_stor_control_thread) from [<8013d99c>] (kthread+0x13c/0x16c)
[77289.129576] [<8013d99c>] (kthread) from [<8010810c>] (ret_from_fork+0x14/0x28)
[77289.129644] INFO: task jbd2/sda1-8:644 blocked for more than 120 seconds.
[77289.129652]       Tainted: G         C      4.14.62-v7+ #1134
[77289.129657] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[77289.129663] jbd2/sda1-8     D    0   644      2 0x00000000
[77289.129691] [<8079d258>] (__schedule) from [<8079d8c0>] (schedule+0x50/0xa8)
[77289.129711] [<8079d8c0>] (schedule) from [<8014b284>] (io_schedule+0x20/0x40)
[77289.129730] [<8014b284>] (io_schedule) from [<8079e230>] (bit_wait_io+0x1c/0x70)
[77289.129749] [<8079e230>] (bit_wait_io) from [<8079dec0>] (__wait_on_bit+0x94/0xd0)
[77289.129768] [<8079dec0>] (__wait_on_bit) from [<8079df84>] (out_of_line_wait_on_bit+0x88/0x90)
[77289.129792] [<8079df84>] (out_of_line_wait_on_bit) from [<802c2eb8>] (__wait_on_buffer+0x40/0x44)
[77289.129818] [<802c2eb8>] (__wait_on_buffer) from [<8037b24c>] (jbd2_journal_commit_transaction+0xdb0/0x17f0)
[77289.129842] [<8037b24c>] (jbd2_journal_commit_transaction) from [<80380d9c>] (kjournald2+0x110/0x2c0)
[77289.129863] [<80380d9c>] (kjournald2) from [<8013d99c>] (kthread+0x13c/0x16c)
[77289.129883] [<8013d99c>] (kthread) from [<8010810c>] (ret_from_fork+0x14/0x28)

I run fsck -fcv on the disk and it found no errors and no bad blocks.