7,435 reputation
621
bio website
location
age 35
visits member for 3 years, 7 months
seen yesterday

1d
comment What is the difference between “inode size” and “Bytes per inode”
Actually extended attributes are stored in their own block rather than the inode, so you don't need a larger inode for them. There were some patches floating around on the mailing lists rather recently to finally add the ability to store extended attributes in the inode if they are small enough to fit, but if they have hit mainline yet, it would have been very recently.
2d
comment Cannot mount external hard drive after using Clonezilla
Looks like the drive is dead. Check dmesg or /var/log/syslog for more detailed error messages.
2d
comment Creating EFI bootable GPT partition with gdisk on previous MBR, GPT Damaged
Did you try the verify or repair commands in gdisk? Don't forget to write the changes after making repairs.
2d
comment Create a binary input file from hdparm's --read-sector option
The man page is a bit misleading then. It doesn't do anything magic. It only crafts the command and sends it down itself instead of letting the kernel do it. The actual command the drive sees is exactly the same.
2d
comment Create a binary input file from hdparm's --read-sector option
What? hdparm --read-sector isn't going to have any easier time reading a bad sector than anything else. You want ddrescue, which can at least skip the bad block and come back to it later, retrying until it hopefully works.
2d
comment Can the hard drive timeout be disabled in Linux (attempting task abort)
Don't have it emulate a scsi disk?
2d
comment /proc/self/maps - 3rd mapped piece of file?
@newbie, that is pretty fscked up. The .init_array and .fini_array should definitely be read only, as they are actually mapped. Not sure why the headers says they aren't.
Dec
15
comment Can the hard drive timeout be disabled in Linux (attempting task abort)
Those error messages do not look like what you describe. Normally a vm has an emulated virtio or ide disk, but those messages appear to be coming from the mpt scsi controller driver.
Dec
15
comment Fdisk cannot read size of this disk is 4.0 TB under RAID-0 enviroment
You aren't making sense; the disk is already partitioned with GPT and so it should be fine. The only problem is that you are trying to look at it with fdisk, and it is telling you that it can't report it correctly ( because you are using a rather old fdisk ). Use parted print and it should show up correctly.
Dec
15
comment Alternatives to LVM
With LVM, you don't lose data except what is on the failed drives by default. This is because by default, the allocation policy is linear, aka JBOD in raid terms. You only lose everything when one drive fails with raid0.
Dec
15
comment Recovering mismatched sectors in a RAID array
@Wyzard, drives don't just suddenly and silently corrupt the data on them. The mismatch is either caused by creating the array initially with --assume-clean when they were not actually clean, or power loss or crash in the middle of a write, in which case, there isn't much point in trying to recover some small part of some file that didn't finish being written.
Dec
15
comment Recovering mismatched sectors in a RAID array
@Wyzard, I suppose, but then there are several alternates, not just one, and there's no real way to find and inspect them, and the correct one is almost certainly just the regular data and it's the checksum that is wrong.
Dec
13
comment take the DRAM performance of the processes into consideration when scheduling
@BeileiSun, yes, and if it is, then whatever task is heavily using it will also be heavily using the cpu, which the scheduler takes into account.
Dec
13
comment /proc/self/maps - 3rd mapped piece of file?
Based on the observed behavior, the reverse is true: it uses the sections header. Only paying attention to the segments table works since the only error it introduces is incorrectly granting execute permission to a few sections that don't need it.
Dec
13
comment /proc/self/maps - 3rd mapped piece of file?
@newbie, because the Sections: header says it is divided into the .got.plt, .data, and .bss segments.
Dec
12
comment different ways how Linux TCP/IP stack terminates connection
There are no SYN flags in the trace you listed; only FIN and ACK. It looks like you simply did not capture the opening of the connection and only got the perfectly normal close.
Dec
12
comment /proc/self/maps - 3rd mapped piece of file?
@newbie, again, it can't since the second mapping is read only, and the data segment must be writable.
Dec
12
comment take the DRAM performance of the processes into consideration when scheduling
@BeileiSun, (n)either. It doesn't really matter whether one is using only the cpu and the other is using the cpu and ram; both need the cpu so that is the only thing that matters to the scheduler.
Dec
12
comment take the DRAM performance of the processes into consideration when scheduling
@BeileiSun, heavy cpu usage does not imply heavy ram usage, but the reverse is true. This is because the cpu can be busy with things other than accessing ram, but you can not be busily accessing ram without using the cpu.
Dec
12
comment /proc/self/maps - 3rd mapped piece of file?
The second mapping can not be .data and .bss since it is read only and those segments are read/write. @Giles was correct here. Use objdump -x to see the different segments and you should find the .data segment lines up with that third mapping.