• The forum software that supports hummy.tv has been upgraded to XenForo 2.1!

    This upgrade brings a number of improvements including the ability to bookmark posts to come back to later. Please bear with us as we continue to tweak things and open a new thread for any questions, issues or suggestions in Site/Forum Issues.

Guidance on using Fixdisk for LBA 0 errors

rodp

Member
Hi All,

After getting some errors reports on the HDD through Webif (and RS) I've been reading around the various threads which talk about fixdisk and LBA 0 errors but still a little unsure in the process and how to fix it. I went into Maintenance Mode and ran fixdisk with no extra commands. The log below shows that it came up with an LBA 0 error. However I chose to not fix it as I wanted to see if there were any other errors. I think the log below shows that there weren't any, unless you understand it better than I?

From the Diagnostic table you can see that I got my first LBA 0 error sometime ago. And after a little discussion on this site I decided it would be best to save everything off my HDD and then reformat it. But I haven't quite got round to it! :(. However, do I need to or should I trust the fixdisk to fix LBA 0? The reason being is I'm aware that LBA 0 is where the partition info resides.

Another question, my Sysmon chart for the HDD is showing that there are 8 sectors (?) offline. As fix disk didn't find any other errors as per below log, are these all in LBA0 then? How many sectors at LBA0 can go wrong before one starts to worry?

My next move is to go back into maintenance mode and re-run fixdisk but say Y to the question about fixing it. I've read that some people had difficulties and there was mention about this 'hdparm' program. Do I need this for fixdisk to do it's stuff or do I need it when fixing sectors manually?

Thanks

Rodp

1572211157449.png

HDD stats - given it's age, it's not doing too bad to only have a count of 24 sectors reallocated. But one thing I don't understand is that is the chart below showing that of those 24 that were reallocated 8 of them have now gone offline again? or should that offline be a step change in the chart?
1572211227242.png

1572211275107.png

1572211295498.png

fixdisk log sorry in reverse order!

Code:
141  
140    ----------------------------------------------------------------------
139    Thu Oct 24 01:47:13 BST 2019: Fix-disk run ended.
138    ----------------------------------------------------------------------
137  
136    Removing extra swap space.
135    Thu Oct 24 01:47:11 BST 2019
134    I/O read: 15633MB, write: 1MB, rate: 3.51MB/s
133    Memory used: 1308k/0k (547k/762k), time: 4456.44/2732.79/185.01
132           13086 files
131    ------------
130               0 sockets
129            1360 symbolic links (1358 fast symbolic links)
128               2 links
127               0 fifos
126               0 block device files
125               2 character device files
124             539 directories
123           11183 regular files
122  
121              87 large files
120               0 bad blocks
119       218747757 blocks used (90.65%, out of 241304332)
118                 # of inodes with ind/dind/tind blocks: 3016/1957/5
117              26 non-contiguous directories (0.2%)
116            1934 non-contiguous files (14.8%)
115           13093 inodes used (0.02%, out of 60334080)
114  
113    Pass 5: I/O read: 58MB, write: 0MB, rate: 0.23MB/s
112    Pass 5: Checking group summary information
111    Pass 4: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
110    Pass 4: Checking reference counts
109    Pass 3: I/O read: 1MB, write: 0MB, rate: 13.93MB/s
108    Pass 3: Memory used: 1308k/18864k (658k/651k), time:  0.07/ 0.05/ 0.00
107    Pass 3A: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
106    Pass 3A: Memory used: 1308k/18864k (678k/631k), time:  0.00/ 0.00/ 0.00
105    Peak memory: Memory used: 1308k/18864k (665k/644k), time: 4114.52/2453.27/182.61
104    Pass 3: Checking directory connectivity
103    Pass 2: I/O read: 6MB, write: 0MB, rate: 1.69MB/s
102    Pass 2: Checking directory structure
101    Pass 1: I/O read: 15568MB, write: 0MB, rate: 3.83MB/s
100    Pass 1: Checking inodes, blocks, and sizes
99    Thu Oct 24 00:32:54 BST 2019: Checking partition /dev/sda2...
98  
97    Creating swap file...
96  
95    Thu Oct 24 00:32:10 BST 2019
94    I/O read: 17MB, write: 1MB, rate: 11.93MB/s
93    Memory used: 140k/0k (50k/91k), time:  1.43/ 0.69/ 0.43
92               6 files
91    ------------
90               0 sockets
89               0 symbolic links (0 fast symbolic links)
88               0 links
87               0 fifos
86               0 block device files
85               0 character device files
84               3 directories
83               3 regular files
82  
81               1 large file
80               0 bad blocks
79           17479 blocks used (6.64%, out of 263064)
78                 # of inodes with ind/dind/tind blocks: 3/3/0
77               0 non-contiguous directories (0.0%)
76               1 non-contiguous file (6.7%)
75              15 inodes used (0.02%, out of 65808)
74  
73    Pass 5: I/O read: 1MB, write: 0MB, rate: 2.85MB/s
72    Pass 5: Checking group summary information
71    Pass 4: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
70    Pass 4: Checking reference counts
69    Pass 3: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
68    Pass 3: Memory used: 140k/0k (71k/70k), time:  0.00/ 0.00/ 0.00
67    Pass 3A: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
66    Pass 3A: Memory used: 140k/0k (72k/69k), time:  0.00/ 0.00/ 0.00
65    Peak memory: Memory used: 140k/0k (72k/69k), time:  0.96/ 0.28/ 0.42
64    Pass 3: Checking directory connectivity
63    Pass 2: I/O read: 1MB, write: 0MB, rate: 81.67MB/s
62    Pass 2: Memory used: 140k/0k (72k/69k), time:  0.01/ 0.00/ 0.00
61    Pass 2: Checking directory structure
60    Pass 1: I/O read: 17MB, write: 0MB, rate: 21.46MB/s
59    Pass 1: Checking inodes, blocks, and sizes
58    Thu Oct 24 00:32:09 BST 2019: Checking partition /dev/sda1...
57  
56    Thu Oct 24 00:32:09 BST 2019
55    I/O read: 166MB, write: 1MB, rate: 13.26MB/s
54    Memory used: 340k/0k (53k/288k), time: 12.52/ 6.51/ 2.45
53               8 files
52    ------------
51               0 sockets
50               0 symbolic links (0 fast symbolic links)
49               0 links
48               0 fifos
47               0 block device files
46               0 character device files
45               2 directories
44               6 regular files
43  
42               1 large file
41               0 bad blocks
40          311480 blocks used (11.88%, out of 2622611)
39                 # of inodes with ind/dind/tind blocks: 4/2/0
38               0 non-contiguous directories (0.0%)
37               1 non-contiguous file (5.9%)
36              17 inodes used (0.00%, out of 655776)
35  
34    Pass 5: I/O read: 1MB, write: 0MB, rate: 0.39MB/s
33    Pass 5: Checking group summary information
32    Pass 4: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
31    Pass 4: Checking reference counts
30    Pass 3: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
29    Pass 3: Memory used: 340k/0k (259k/82k), time:  0.00/ 0.00/ 0.00
28    Pass 3A: I/O read: 0MB, write: 0MB, rate: 0.00MB/s
27    Pass 3A: Memory used: 340k/0k (259k/82k), time:  0.00/ 0.00/ 0.00
26    Peak memory: Memory used: 340k/0k (260k/81k), time:  8.93/ 3.57/ 2.41
25    Pass 3: Checking directory connectivity
24    Pass 2: I/O read: 1MB, write: 0MB, rate: 119.37MB/s
23    Pass 2: Checking directory structure
22    Pass 1: I/O read: 164MB, write: 0MB, rate: 19.41MB/s
21    Pass 1: Checking inodes, blocks, and sizes
20    Thu Oct 24 00:31:56 BST 2019: Checking partition /dev/sda3...
19  
18    Using superblock 0 on sda3
17    Using superblock 0 on sda2
16    Using superblock 0 on sda1
15  
14    GPT Status: not present
13    MBR Status: MBR only
12  
11    Checking partition tables...
10    Do you wish to attempt repair of the bad block? [Y/N]: Skipped repair of LBA 0
9    Error at LBA 0
8    Running short disk self test
7  
6    Checking disk sda (4096 byte sectors)
5  
4    ----------------------------------------------------------------------
3    Thu Oct 24 00:29:10 BST 2019: Fix-disk run starting...
2    ----------------------------------------------------------------------
1
 
Last edited:

MartinLiddle

Super Moderator
Staff member
The log below shows that it came up with an LBA 0 error. However I chose to not fix it as I wanted to see if there were any other errors. I think the log below shows that there weren't any, unless you understand it better than I?
I would login via telnet and use gfdisk to list the partition table and save the output. The partition table can be rebuilt if necessary but I think it is probably wise to know what it should look like.

Another question, my Sysmon chart for the HDD is showing that there are 8 sectors (?) offline. As fix disk didn't find any other errors as per below log, are these all in LBA0 then? How many sectors at LBA0 can go wrong before one starts to worry?
I am guessing this is a disk with 4K sectors so all reports will be in multiples of 8. I don't think it is a major concern with only 24 reallocated sectors.
My next move is to go back into maintenance mode and re-run fixdisk but say Y to the question about fixing it.
That is what I would do.
 
OP
R

rodp

Member
Thanks for the reply. I'll go with running fixdisk (I'll backup first! ;) )

use gfdisk to list the partition table
oooo, I didn't know you could use something called gfdisk to save the partition info. Obviously save it off the drive onto something else?! ;) Just had a quick search for info on it but there doesn't seem to be much. Can you provide some info on saving partition info (and then loading it)?

Thanks

Rodp
 

prpr

Well-Known Member
After getting some errors reports on the HDD through Webif
It amuses me that you obscured certain info. in the inline image, and then included the whole unobscured thing as an attachment.
unless you understand it better than I?
"... better than me".
I decided it would be best to save everything off my HDD and then reformat it.
Backing up is probably a good idea, but what do you expect reformatting to achieve? Technically, reformatting just recreates filesystems and there's nothing wrong with them. Although some people (and software) also mis-use it to mean re-partitioning as well, hence the confusion.
It depends on how exactly you were planning on doing this.
As fix disk didn't find any other errors as per below log, are these all in LBA0 then?
No, they are LBA 0-7.
How many sectors at LBA0 can go wrong before one starts to worry?
None.
one thing I don't understand is that is the chart below showing that of those 24 that were reallocated 8 of them have now gone offline again?
No, it's a different 8. When they are fixed, the reallocated total will go up to 32 and the pending/offline counts should be 0.
I didn't know you could use something called gfdisk to save the partition info.
Just use normal fdisk and make a note of the start/end numbers for the 3 partitions: fdisk -lu /dev/sda
I don't entirely trust gfdisk with its strange calculation methods (although it doesn't (can't) screw up start/end numbers) and its ridiculous obsession with cylinders (an archaic concept).
 

/df

Active Member
...
My next move is to go back into maintenance mode and re-run fixdisk but say Y to the question about fixing it. I've read that some people had difficulties and there was mention about this 'hdparm' program. Do I need this for fixdisk to do it's stuff or do I need it when fixing sectors manually?
...
An up-to-date-ish (2016) version of hdparm is part of the customised firmware. fixdisk -y is supposed to use it to force the disk to relocate any bad sector found by SMART testing; it will then tell you which (if any) file was affected.
... Can you provide some info on saving partition info (and then loading it)?
fixdisk does know about the likely partitioning of a disk in the HDR so that it can generally reconstruct a bad partition table, but there's no harm in keeping a note, or even a restorable copy. You can use the flash filesystems (where the CFW programs are) to store a small file like a partition list, or a USB stick (sometimes known as UPD here).

sfdisk, used by fixdisk, produces better output than fdisk or gfdisk.
Code:
sfdisk --dump /dev/sda >/var/lib/humaxtv_backup/partition/rodspart.dmp
# to restore:
# sfdisk /dev/sda </var/lib/humaxtv_backup/partition/rodspart.dmp
 
Last edited:
OP
R

rodp

Member
@/df,

Thanks for the info.

@prpr,

It amuses me that you obscured certain info.
Thought that clicking undo would have removed the pasted object after realising I needed to change it. Oh well you sometimes find out the hard way! Have edited it. Thanks for flagging it. :)

Rodp
 

Trev

The Dumb One
"...better than I am." In your version, there is a comparison without a verb, so it must be assumed.
Once you stick in the verb, it's blindingly obvious.
 

/df

Active Member
There's no blinding certainty about "than me/I", as with so many grammatical disputes.

If you think that you are using "than" as a conjunction but omitting the verb from the comparative clause that it introduces, as Trev does, then you are right to say "... than I".

If you think that you are using "than" as a preposition with a strong or disjunctive form ("me", cf French "moi") of the 1st person singular pronoun, as prpr may do, then you are right to say "... than me".

The standard introductory work "A student's introduction to English grammar" by Pullum and Huddleston (2005), based on their comprehensive Cambridge Grammar of the English Language, says that both forms are grammatical but "than I" is preferred in very formal contexts. In those contexts you should perhaps not be omitting the verb anyway.

Personally I find it a stretch to believe that whenever I use "<comparative adjective/adverb> than <whatever> <end of phrase>" I meant to use a verb but omitted it.
 
Top