Hyper Backup error message: Your disk has I/O error

[ruckb]

Hello,

I hope somebody can help me with my problem
NAS: DS1019+
DSM: DSM 7.1-42661 Update 3

I make with Hpyerbackup 2 local backups on 2 Seagate Backup Hub USB harddisks (ntfs formated still 5GB free).
This worked for a long time, but now I get a error Messges when I try to start the backup.
The most important:
Failed to access the backup destination. (Your disk has I/O error. Please refer to What should I do if the message "I/O error" appears when using Hyper Backup to back up tasks to local destinations? - Synology Knowledge Center)

I got this on both Harddisk in very short temporal distance (13.07 vs. 19.07 last working Backup)
of course can be a coindicence ...

I can access both backups on the diskstation as well as with Hyperbackup Explorer on a Windows PC

Acoccoring to the link to synology above I did a checkdisk on one HDD and this really resulted in some fails (sorry in german):
"Die zugeordnete Länge 0x7e3000 ist kein Mehrfaches von 0x10000 für das Attribut vom Typ 0x80 und der Instanzkennung 0x2."
After repair the error was gone.
SMART Test from Seatgate Seatools was PASS (sorry, this one only reported PASS, but no detailed report)
But now I get th error message:
Exception occurred while backing up data. (The data stored in backup destination [Backup] was corrupted. Please contact Synology Technical Support.)

I still can access the backup on the diskstation...

In the meantime I formated this one and a new backup is just running, so far this seems to work.

But I would not like to loose the backup from old timestemps, so I do not want to format the second HArddisk as well..

any Idea how I can fix this issue with out deleting the backup on the second HDD?

Best Regards
Hermann

 

[pbondurant]

As per the synology link you posted, you need to read the SMART data on the USB drives:

"After the test is complete, please refer to ID fields 5, 197, and 198. For a new USB drive or a USB drive in normal operation, the Raw Data values in these fields should be 0. If the Raw Data values are not 0, it means that issues may have occurred to the USB drive."
This will confirm whether the HDDs likely have a hardware problem or not. A utility like smartctl will read these values for you.

The error message from Hyperbackup ("Exception occurred while backing up data. (The data stored in backup destination [Backup] was corrupted. Please contact Synology Technical Support.)") is a generic error that can be caused by either hardware problems or problems with Hyperbackup itself.

IME when I ran a Synology, I had frequent failures with Hyperbackup. It would run ok for a few weeks/months, and then I'd get the error you mention above. HDD passed all SMART checks; my only option was to wipe the drive and restart teh Hyperbackups from scratch. I always suspected that Hyperbackup was junk and had bugs that would appear over time possibly due to multiple versions of files being backed up. There are lots of reports of this error on teh internet without any definite fix.

Because of this, I moved to a 3rd party backup (restic) and never had problems again, and using the same external drives as before.

hth

 

[ruckb]

hi pbondurant,
thanks for the feedback.
as seatools for windows does only result in a PASS, but not in any real result I guess I would need to use seatools for DOS. I expect other tools (e. g. SMARTMONTOOLS) will not easily interprete results correctly (espeically for USB HDD).

So far I was happy with Hyperbackup... it worked for several years without issues.
The nice thing was: after Backup the HDD is ejected automatically, in combination with a programmable power plug I have the HDD only connected during the backup is made, and do not need to conect anyting manally.
This was causing a bit of headacke when I initially setup the backup process.
how would you eject the HDD wtih restic automatically?
(i tried with independent task, but could not set it up corretly :-( until I figured out, that Hyperdisk does this automatically..)

Hermann

 

[pbondurant]

Smartmontools will report smart data from most common hdds, including USB attached ones. The only way to know if it will work with your particular HDD is to try!

Restic is a command line backup program. If you want to eject a drive after a backup has run you would do this trivially as part of a backup script, eg

sudo umount /media/<hdd_name>> && sudo udisksctl power-off --block-device /dev/sda

...but this is linux / macos stuff. I'm not familiar with backup options using a windows machine.

EDIT: It seems restic is available for windows too restic · Backups done right! . But it sounds like maybe you're looking for something less technical & more automated/gui.

 

[ruckb]

Hello pbondurant,

SMART: especially for USB HDD i got very rare usefull results by different SMART tools (including smartmontools). But i did not try with my hard disk so far and the trials I'm talking about are somehow long ago.
Usually the best restults that I got have been with the tools of the Hard disk vendor..
need to give diffent toptions a try (once I find some time..)


Thanks for the unmount script ... I found several different scripts for unmounting the HDD from the NAS, but so far did not get them running .. will give yours a trial.

I checked out restic, and it looks interesting. In the past i did script based backup .. but i do not have the time any more too adjust all things manually (I am too old .. time is running out ;-) ) things need just to work out of the box, so i can concentrate on my real work...
So you are right .... I look to something more automated/gui like, but it should run on the Diskstation..

Hermann

 

[EAZ1964]

Disk I/O errors are often caused by bad connections.
In those cases you will not find an issue using smart, checkdisk or seatools.
Your cable or connector may be worn out, broken or dirty.

 

[pbondurant]

Disk I/O errors are often caused by bad connections.
In those cases you will not find an issue using smart, checkdisk or seatools.
Your cable or connector may be worn out, broken or dirty.

UDMA CRC Error Count (Smart id199) is often a good indicator of a bad sata cable. This value checks for data inconsistencies that occur between the motherboard sata controller <-> HDD controller and a bad cable/connectors is usually the culprit if this count steadily increases.

 

[ruckb]

Smartmontools worked ..

5, 197, 198,199 have raw data "0" so i expect no problem (see data below).
but I do not like the Raw Read Error Rate, seek error rate and Hardware ECC Recovered numbers..


For id 199: I expect this will not help for bad USB cables (for native SATA drives/cables I fully agree, that this can be an indication for cabling issues)
I expect that the USB communication finds bad USB cables before the SATA controller will get any information about any transmission fails.

What is still find strange:
eEven checkDisk says "no errors found" the Details show still the following messages:
"Die zugeordnete Länge 0x7f1000 ist kein Mehrfaches von 0x10000 für das Attribut vom Typ 0x80 und der Instanzkennung 0x2."
DeepL translation:
"The allocated length 0x7f1000 is not a multiple of 0x10000 for the attribute of type 0x80 and instance identifier 0x2."

Can this be the reason why Hyperbackup says the bakup is corrupted when trying to run the backup task?
(as mentioned: Hyperbackup can access the backup on the diskstation and on my windows PC)

Hermann



=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAGS VALUE WORST THRESH FAIL RAW_VALUE
1 Raw_Read_Error_Rate POSR-- 072 064 044 - 15787475
3 Spin_Up_Time PO---- 089 089 000 - 0
4 Start_Stop_Count -O--CK 100 100 020 - 406
5 Reallocated_Sector_Ct PO--CK 100 100 010 - 0
7 Seek_Error_Rate POSR-- 081 068 045 - 130998647
9 Power_On_Hours -O--CK 094 094 000 - 5871 (19 161 0)
10 Spin_Retry_Count PO--C- 100 100 097 - 0
12 Power_Cycle_Count -O--CK 100 100 020 - 405
184 End-to-End_Error -O--CK 100 100 099 - 0
187 Reported_Uncorrect -O--CK 100 100 000 - 0
188 Command_Timeout -O--CK 100 100 000 - 0 0 0
189 High_Fly_Writes -O-RCK 080 080 000 - 20
190 Airflow_Temperature_Cel -O---K 068 044 040 - 32 (Min/Max 30/32)
191 G-Sense_Error_Rate -O--CK 099 099 000 - 3699
192 Power-Off_Retract_Count -O--CK 100 100 000 - 46
193 Load_Cycle_Count -O--CK 100 100 000 - 1096
194 Temperature_Celsius -O---K 032 056 000 - 32 (0 23 0 0 0)
195 Hardware_ECC_Recovered -O-RC- 004 001 000 - 15787475
197 Current_Pending_Sector -O--C- 100 100 000 - 0
198 Offline_Uncorrectable ----C- 100 100 000 - 0
199 UDMA_CRC_Error_Count -OSRCK 200 200 000 - 0
200 Pressure_Limit PO---K 100 100 001 - 0
240 Head_Flying_Hours ------ 100 253 000 - 416h+31m+15.171s
241 Total_LBAs_Written ------ 100 253 000 - 58704876417
242 Total_LBAs_Read ------ 100 253 000 - 8359391131
||||||_ K auto-keep
|||||__ C event count
||||___ R error rate
|||____ S speed/performance
||_____ O updated online
|______ P prefailure warning

 

[pbondurant]

Smartmontools worked ..

5, 197, 198,199 have raw data "0" so i expect no problem (see data below).
but I do not like the Raw Read Error Rate, seek error rate and Hardware ECC Recovered numbers..


For id 199: I expect this will not help for bad USB cables (for native SATA drives/cables I fully agree, that this can be an indication for cabling issues)
I expect that the USB communication finds bad USB cables before the SATA controller will get any information about any transmission fails.

What is still find strange:
eEven checkDisk says "no errors found" the Details show still the following messages:
"Die zugeordnete Länge 0x7f1000 ist kein Mehrfaches von 0x10000 für das Attribut vom Typ 0x80 und der Instanzkennung 0x2."
DeepL translation:
"The allocated length 0x7f1000 is not a multiple of 0x10000 for the attribute of type 0x80 and instance identifier 0x2."

Can this be the reason why Hyperbackup says the bakup is corrupted when trying to run the backup task?
(as mentioned: Hyperbackup can access the backup on the diskstation and on my windows PC)

Hermann



=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAGS VALUE WORST THRESH FAIL RAW_VALUE
1 Raw_Read_Error_Rate POSR-- 072 064 044 - 15787475
3 Spin_Up_Time PO---- 089 089 000 - 0
4 Start_Stop_Count -O--CK 100 100 020 - 406
5 Reallocated_Sector_Ct PO--CK 100 100 010 - 0
7 Seek_Error_Rate POSR-- 081 068 045 - 130998647
9 Power_On_Hours -O--CK 094 094 000 - 5871 (19 161 0)
10 Spin_Retry_Count PO--C- 100 100 097 - 0
12 Power_Cycle_Count -O--CK 100 100 020 - 405
184 End-to-End_Error -O--CK 100 100 099 - 0
187 Reported_Uncorrect -O--CK 100 100 000 - 0
188 Command_Timeout -O--CK 100 100 000 - 0 0 0
189 High_Fly_Writes -O-RCK 080 080 000 - 20
190 Airflow_Temperature_Cel -O---K 068 044 040 - 32 (Min/Max 30/32)
191 G-Sense_Error_Rate -O--CK 099 099 000 - 3699
192 Power-Off_Retract_Count -O--CK 100 100 000 - 46
193 Load_Cycle_Count -O--CK 100 100 000 - 1096
194 Temperature_Celsius -O---K 032 056 000 - 32 (0 23 0 0 0)
195 Hardware_ECC_Recovered -O-RC- 004 001 000 - 15787475
197 Current_Pending_Sector -O--C- 100 100 000 - 0
198 Offline_Uncorrectable ----C- 100 100 000 - 0
199 UDMA_CRC_Error_Count -OSRCK 200 200 000 - 0
200 Pressure_Limit PO---K 100 100 001 - 0
240 Head_Flying_Hours ------ 100 253 000 - 416h+31m+15.171s
241 Total_LBAs_Written ------ 100 253 000 - 58704876417
242 Total_LBAs_Read ------ 100 253 000 - 8359391131
||||||_ K auto-keep
|||||__ C event count
||||___ R error rate
|||____ S speed/performance
||_____ O updated online
|______ P prefailure warning

195 Hardware_ECC_Recovered is a measure of time between errors, so a high value is good: Should I be concerned about a high SMART Hardware_ECC_Recovered value?

1 Raw_Read_Error_Rate - there are always some raw read errors here; teh important value is the '72' in the 'Value' column. 100 is perfect, 1 is bad; 72 suggests some raw read errors that the firmware has corrected, but nothing too extreme. It wouldn't cause me to replace the drive at this value Understanding SMART Reports - Unraid | Docs

7 Seek_error_rate - see following for an explanation. With Seagate drives, the raw value has to be decoded; its not a simple decimal value:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

The HB error is simply telling you that HB thinks some aspect of the previous backups are corrupted, ie its a corruption of the HB data structures rather than a hardware error.
Try googling it but IME these HB specific errors are often not well documented online. You could try Synology support ofc.
To repeat, I had many such HB errors reporting data corruption over the years, but the hdds always checked out fine. The only solution was to wipe and recreate the backups which would run ok for a few months until another failure. IMO the HB app is buggy and unreliable. It works ok if you just need to have a few recent backups of everything in case of a disaster; ime it's not really fit for long term backup archival use.

hth

 

[ruckb]

thanks for the good explanations!
I'm in contatct with synology, but so far it does not look like they will have any real good answers.
currently the proposal is to "backup my backup" to keep the old versions, wipe the disk and restart the backup ..
This is what i have done with my first HDD already, and i guess now will do with the second one as well.


Hermann