Recently purchased a Toshiba Satellite C50PSCJEA-01N011 Celeron-N2820 4G
500GB 15.6" Windows8.1 Notebook PSCJEA-01N011
Sad reality is because a certain piece of medical equipment requires 'doze
(no it doesn't work with Wine etc) said machine had to be bought with said
Obviously don't want to do much in the MS-Windows world and have sought to
navigate my way around the madness that is MS-Windows 8; have resized the
disk, turned off secure boot (necessary for installation apparently),
began installation with a nice Debian Mint 201403 disk, grub comes up
I get a black screen. The DVD drive happily whirls around obviously
wanting to entertain the possibility of another Linux install, but alas
with no screen display not much else can be done.
I have attempted a standard GRUB modification used on other Toshiba
systems (https://coderwall.com/p/ydbldg) but to no avail.
Lev Lafayette, BA (Hons), GradCertTerAdEd (Murdoch), GradCertPM, MBA (Tech
mobile: 0432 255 208
RFC 1855 Netiquette Guidelines
I have been reading the "btrfs/ZFS, sans raid and bitrot" thread and a
number of thoughts and questions spring to mind.
I get the impression that some are looking for a single reliable storage
solution to avoid having to do backups.
Surely this is impossible, I certainly would ______NEVER______ (excuse the
shouting) ever trust my life to a single system if at all possible. When
one is doing instrument flying training as a pilot you are constantly told
never to rely on a single instrument but scan all of them and come up with
an overall coherent picture. If one relies in such circumstances on a
single point of failure you __will__ kill yourself.
One is told raid or any such thing is a reliabilty strategy __not__ a
I personally keep all data I consider important on four separate
systems/devices one device (which is in fact duplicate items but differing
technolgies) being kept off site. Maintaining this is a bit of a pain but
there is no other way as far as I can see.
The reason for the number of separate backups is we had in one instance in
a large commercial situation managed to destroy two backs trying to restore
a system. We only succeeded in the end becuase I had independantly
duplicated one of the backups on another system.
After reading about bitrot and feeling guilty for storing my most valuable
data on cheap drives (although with backups!) I've been thinking about
moving to something more resilient.
My current setup is a Ubuntu laptop, with 2 external drives.
1X 2TB ext4 for data storage
1X 3TB ext4 for backup (using Crashplan commercial backup software).
My question, is if I change the first drive to btrfs or ZFS, will I gain
resiliency from bitrot?
My understanding is I need 2 drives in at least a RAID 1 to get automatic
healing from bitrot, but if I at least use a filesystem with check summing
support then I will be able to at least restore my affected files from my
Crashplan backups (which are compressed then checksummed and regularly
checked for errors automatically) and I won't have the risk of my main
drive corrupting my backups, because the read will FAIL if it doesn't pass
Is my understanding correct?
Here is a list of some things that I think would be good for the hardware
library, if you have something that matches which you don't need then please
bring it to a LUV meeting to donate it. Also just bring random stuff that's
small and light.
DDR2 and DDR3 RAM - I think there is plenty of older RAM in the library
Laptop RAM. Not many people want it but those who do want it badly. It also
works in lots of modern printers.
USB Flash storage. While most of us get enough freebies from trade shows I'm
sure that some people need more. It's small and light so I don't mind
carrying them around.
64bit PCs. I'm not planning to carry spare machines to regular meetings, but
I'm happy to store a few in case someone needs one and bring them to a meeting
on demand. Post to the list first if you plan to bring a PC to the meeting.
ATI PCIe video cards. NVidia has driver issues (either a lack of features or
non-free depending on which driver you choose).
Small Ethernet switches that are quiet. Even a 10baseT switch is very handy
if it's small and light.
Micro-USB cables suitable for connecting Android devices (and many other
things) to PCs.
USB power supply as used for phones and tablets.
Android devices. Even devices that run Android 2.1 can be useful if they
aren't broken. More recent Android devices can be useful even if they are a
bit broken (EG a cracked screen still leaves an Android device as a portable
computer with a camera).
Accessories for phones (please clearly label which phone they belong to).
There are multiple standards for things like external microphones for phones,
some old phones are unreasonably expensive to buy parts for. Batteries are
useful if they can take at least 50% original charge, please wrap batteries in
plastic to avoid the risk of short-circuiting.
Here are some things that I personally need:
A US style extension cable that goes from a PC PSU to a monitor. It's not a
very important thing for me but would be nice to have. As few PCs have the
socket for such a cable nowadays there should be a few unused out there.
A quad-core PC with 4*DDR2 sockets that can take full height PCIe cards.
SATA disks with many bad sectors. The ideal would be to have about 0.1% of
reads fail. I'm doing some research into error recovery. Please clearly
write "BAD" on such disks to avoid unfortunate mistakes. Note that when I am
finished experimenting with such disks I will use a hammer to make them
unreadable - just in case the disk problems made it impossible to delete some
of your data.
My Main Blog http://etbe.coker.com.au/
My Documents Blog http://doc.coker.com.au/
I'm working on writing /etc/magic entries for ZFS. Below is the start that
I've made. It uses XDR for the header which has pascal strings with a 4 byte
big-endian length. I thought that should be "pstring/L" but the following
pstring types are the ones that worked.
Also running "file -s -k /dev/sd?" doesn't work, is the -k option expected to
work on a device node?
0x4010 belong 0x20
>0x4014 pstring/l version ZFS store version
>>0x4020 belong 8
>>>0x4024 belong x %d
>>>0x403a pstring/h name
>>>>0x4048 pstring/l x pool name %s
My Main Blog http://etbe.coker.com.au/
My Documents Blog http://doc.coker.com.au/
TLDR: I am running out of time to buy a webcam for use with Skype on
Debian, can anyone suggest a brand or model that is known to work?
I'm asking for tips because I am on a tight timeline and the task is 5
hours drive away at a relative's house, where I'm going to install
They have requested Skype, I've not used it nor even webcams before.
I went to the local computer shop and plugged in a webcam and saw
0c45:62c0 in dmesg.
But checking at home the Debian IRC bot "judd" says "Unknown device
from unknown vendor".
So I'm looking for tips to at least purchase something that will work.
(I'm at work now so I cannot test anything.)
Thanks for any assistance
Does anyone have real-world experience of using linux's interface
bonding on public networks?
(In the bandwidth-aggregation mode, not the redundancy mode)
I was wondering how I could make the following setup work:
* Rent a VPS in Melbourne with four IP addresses
* Get four (or just two) ADSL connections wired up to home
* Have your VPS connect four VPN connections from itself back to each
of your home IPs.
* Bond all four interfaces together
* Create a fifth VPN connection, this time going over the
bonded-virtual-interface between VPS and home, and then configure your
home server to use that link as the default route?
It sounds pretty messy and I'm not sure it'd actually work in
practice; the routing tables would be hell to get right.
Are there any guides already out there?
With autofs on Red Hat/Centos, the" /net -hosts" special map allows all NFS
resources exported by all accessible NFS servers to get mounted under the
/net directory without explicitly mounting each one of them. E.g. accessing
/net/host1 will instruct autofs to mount all available resources on host1.
At first, it didn't work. E.g. if I run
I have always got "No such file or directory".
After running tcpdump to analyse the packets, I discovered that this
particular feature uses sunrpc (port 111) and a random UDP port (30000
above). The problem is how I set iptables rule for this random port. I know
I can specify a port range like --dport 30000:60000, but this is not a safe
way to configure a firewall.
Does anyone want a Tenda D820B ADSL 2+ modem?
I have NBN now, so it's surplus. It's a modem only, not a router, so it
needs to be managed by a router running PPPoE to work. I had it running
with a WRT54G running OpenWRT for years - worked fine.
Pickup in Brunswick West
Erik Christiansen <dvalin(a)internode.on.net> said,
>Here in Tecoma (Dandenong Ranges), my ISP (Internode) provides ADSL2+.
>Is it a rural exchange which would limit it to ADSL1, or the 5 km line
Im practice it would be line length. As far as I am aware ADSL uses line
frequecies up to around 1mhz, ADSL 2 to around 2.2 mhz. Special 1.27mm
copper carrier cable has a loss at 1.6mhz of around 30db for 5 kilometres.
One would expect most customer cable would be around double that. Such a
loss would put receive levels for ADSL2 at 5000 metres well down into the
noise (-50 to -60dbm). Even for ADSL you would need real good copper cable
for it to work at 5000 metres.