Hard Disk Monitoring

Hi

Win10 has reported an issue with a drive. I spotted it in Windows Event Viewer - there were lots of disk and filesystem errors:

Disk:
The IO operation at logical block address 0x5f54a0 for Disk 1 (PDO name: \Device\00000024) was retried.
The IO operation at logical block address 0x5f5058 for Disk 1 (PDO name: \Device\00000024) was retried.

Volsnap
The shadow copies of volume F: were aborted because of an IO failure on volume F:.

So I checked the disk for errors (Win10 File Explorer only has one type of disk check now vs Win7's 2 types). Event Viewer sais:

Ntfs:
Volume F: (\Device\HarddiskVolume4) needs to be taken offline for a short time to perform a Spot Fix. Please run "CHKDSK /SPOTFIX" locally via the command line, or run "REPAIR-VOLUME <drive:>" locally or remotely via PowerShell.

Too many repair events have occurred in a short period of time.
Temporarily suspending posting of further repair events.

Check disk said that it was unable to scan the drive and to repair the drive if this continued.

So I clicked on "Repair drive" - Event Viewer says:

Details:
Chkdsk was executed in scan mode on a volume snapshot.

Checking file system on F:
Volume label is Vol 1 Backup.

Stage 1: Examining basic file system structure ...
Found 0x1 clusters allocated to file "[snip filename]<0x1,0x7c794>" at offset "0" marked as free
... repaired online.
Found 0x1 clusters allocated to file "[snip filename]<0x1,0x7c795>" at offset "0" marked as free
... repaired online.

[snip lots of entries]
File "[snip filename]<0x1,0x7c7b9>" and file "[snip filename]<0x1,0x7c810>" both own logical cluster 0x3b7307f
... queued for offline repair.

etc. etc.

Chkdsk was executed in spotfix mode on a read-write volume.

Checking file system on F:
Volume dismounted. All opened handles to this volume are now invalid.
Volume label is Vol 1 Backup.

Examining 9 corruption records ...

Record 1 of 9: File "[filename-snip] <0x1,0x7c7b4>" and file "[filename-snip]<0x1,0x7c80c>" both own logical cluster 0x3b73073 ... corruption found and fixed.
....
9 corruption records processed in 3.3 seconds.

Windows has fixed all previously identified issues with this drive.
No further action is required.
Write failure with status 0xc0000483 at offset 0x5e8982000 for 0x19400 bytes.
Write failure with status 0xc0000483 at offset 0x5e8982000 for 0x19400 bytes.

By contrast HDS says:
Drive health 100%

Overview says:
Problems occurred between the communication of the disk and the host 59174 times.

No actions needed.

Log says:
23/02/2016 16:52:18,#199 Ultra ATA CRC Error Count 59144 -> 59174
23/02/2016 15:22:11,#199 Ultra ATA CRC Error Count 59074 -> 59144
23/02/2016 15:16:53,#199 Ultra ATA CRC Error Count 59011 -> 59074
23/02/2016 10:36:40,#199 Ultra ATA CRC Error Count 59007 -> 59011
22/02/2016 21:55:48,#199 Ultra ATA CRC Error Count 58901 -> 59007
22/02/2016 18:45:09,#199 Ultra ATA CRC Error Count 58888 -> 58901


It confuses me slightly that the large number of problems results in "No actions needed"!

I ran a short self-test anyway which completed successfully.

Any reasons why HDS isn't concerned?

Hi,

As you can see, Hard Disk Sentinel detects and reports numerous data communication problems.
Not sure which Hard Disk Sentinel version do you use, but the latest should display additional detailed information in the text description:

Problems occurred between the communication of the disk and the host 59174 times.
In case of sudden system crash, reboot, blue-screen-of-death, inaccessible file(s)/folder(s), it is recommended to verify data and power cables, connections - and if possible try different cables to prevent further problems.
More information: http://www.hdsentinel.com/hard_disk_cas ... _error.php

Exactly as you can see, data transfer / communication errors can cause file system corruptions, damaged files / folders and so.
Then the logical drive (partition) may become corrupted, chkdsk will happily find and "repair", but this repair is only related to the logical volume / partition.

These issues (even if you may think so) are usually completely independent from the hard disk, but caused by cables, connections.
This is why
- the health is 100% in Hard Disk Sentinel
- the short self test (which runs "inside" the hard disk, so generally without data communication between the hard disk and the computer) runs without problems. Usually the extended self test would be the same, but a Disk menu -> Surface test -> Read test (which transfers data between the hard disk and the computer) may show problems and increased error count.

I'd suggest to check the recommended page: http://www.hdsentinel.com/hard_disk_cas ... _error.php
as it has extensive information about such issue: how it can happen (what can cause it) and how to avoid (by checking connections and replacing cables), diagnose the problem - and then clear the error counter with Hard Disk Sentinel, so then only possible new problems will be detected and reported.

Have you figured out your problem?

Tx for the info. I replaced the SATA cable and the CRC and NTFS errors have stopped.
It would have been good to have some form of pop-up to flag this error though as I only discovered it by chance in the Windows event log.

Generally, the text description and the log of Hard Disk Sentinel should show these kind of problems.
Hard Disk Sentinel can also issue an alert when any such (or other kind) of error detected and added to the error log (see: Configuration -> Alerts -> When a new log entry added. You can associate any alerts (sound, display/send alert, etc.) which will be triggered when this happens).

Yes, I knew that the notifications could be sent for the Log but I had not enabled it as I considered that I might be overrun with daily info-type events, as it's 'just a Log' (speaking from experience on other logging systems!).
However on reading the help, it states that the Log actually only contains "The degradations of the data field of the most important S.M.A.R.T. attributes" so in fact it's more of an "SMART errors you should really know about" log so I now feel more comfortable enabling it for alerts.

Also IIRC the alerts aren't enabled by default for any disk warnings at all, which seems a little odd. Why would one not want to be alerted to disk warnings? One can always disable them afterwards.