Re: [PLUG] Repairing a dead MacBook Pro
I've never looked at a schematic for the Macbook, but if your third HDD is now bad I think there may be a problem with the power supply or an internal voltage regulator that is frying the electronics of your HDD's. On 09/17/2014 02:30 PM, Loren M. Lang wrote: I purchased a 2009-era MacBook Pro a year ago, replaced the bad SATA drive, and promptly put Linux Mint on it. A few months ago, it failed with what looked like a hard drive issue. I booted up with a Linux LiveCD and tried to access the hard drive. dmesg reported that a SATA drive was indeed installed, but unreadable. I tried using hdparm -iI /dev/sda to retrieve the serial number and received an error back trying to read any information. It knew that a SATA drive was there but couldn't seem to acknowledge any commands so it appears that the logic board failed and it was not a mechanical error. I filed an RMA with Western Digital and had my under-warranty hard drive promptly replaced in a couple weeks. I installed the drive only to discover it had the exact same error as the previous drive. I checked the serial number on the label and it was indeed different than the one I sent in for the RMA. As another test, I tried that hard drive in a USB 3.0 enclosure on a separate Linux computer and found that the hard drive was indeed the one spewing those error messages. I then found a third, spare hard drive that I first tested in my USB 3.0 enclosure. It worked so I installed it in my MacBook Pro and booted up my LiveCD yet again. Sure enough, it had the same error and could not even report it's serial number to me. I powered down and moved the spare hard drive back to the enclosure and found that that hard drive is now dead as well. It appears that the real issue is with my MacBook Pro and it's killing my hard drives. I'm assuming the power supply voltages are out of spec so I pulled out my voltmeter. Measuring the SATA power connector with the dead hard drive attached and Mac powered on, I found 5V on the 5V supply lines and no voltage on the 12V or 3.3V lines. No indication of something that might kill a logic board, but I would expect to see 12V on the supply. What's even more odd, if the issue is with the power supply, I would expect it to also affect other SATA devices, at a minimum, but yet, I can completely boot up with a LiveCD on the SATA DVD drive with no issues. I'm at a loss as what to do next to fix my poor MacBook Pro. My Mac friends keep telling me to pull out AppleCare and get it fixed right, but that's a lot of money to fix what's probably just a bad solder joint. ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Accidental X-windows kill
You were using KDE it sounds like, and the key sequence you inadvertently hit was ctrl+alt+esc which is a -really- old X11 shortcut to killing the task that you click on. I know KDE still retains that shortcut, it's one you're typically unlikely to accidentally hit, but it can happen! I don't think Unity kept it though. Anyway, if it happens again, just hit esc to clear it. I actually use it to kill unresponsive steam games or when Firefox decides to stop behaving like an adult. It doesn't just close the window, it terminates the application, so in the example of Firefox, it'll close every Firefox window you have open. It definitely isn't a clean shutdown, but it's good for when an X application has become entirely unresponsive, but you know your system is still good. Pleasantly, Ronald Bynoe From: plug-boun...@lists.pdxlinux.org [plug-boun...@lists.pdxlinux.org] on behalf of Jim Garrison [j...@jhmg.net] Sent: Wednesday, September 17, 2014 10:15 PM To: plug@lists.pdxlinux.org Subject: Re: [PLUG] Accidental X-windows kill On 9/17/2014 10:07 PM, website reader wrote: Has anyone ever hit this before? I was trying to build some special script files for vim, and had to enter a control-v esc character sequence into the script. I tried twice, got a system activity pop-up window which shows active jobs (obviously a desk-top shortcut) and I hit esc again (or control-V or both). Suddenly a dead-man skull icon appears, my mouse goes inactive, I cannot click on any of my X-windows and the skull is the mouse pointer. [snip] The usual way to kill X-Windows is Ctrl-Alt-Backspace, but this is normally disabled by default on modern systems. You didn't say what distro and window manager (Gnome, KDE, etc) you are using, which might be relevant. Or, you might have found a bug in X :-( It would be very interesting if you can reproduce it and figure out exactly what it takes to crash it. -- Jim Garrison (j...@acm.org) PGP Keys at http://www.jhmg.net RSA 0x04B73B7F DH 0x70738D88 ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Accidental X-windows kill
On Wed, Sep 17, 2014 at 10:07 PM, website reader website.read...@gmail.com wrote: Has anyone ever hit this before? I was trying to build some special script files for vim, and had to enter a control-v esc character sequence into the script. I tried twice, got a system activity pop-up window which shows active jobs (obviously a desk-top shortcut) and I hit esc again (or control-V or both). Suddenly a dead-man skull icon appears, my mouse goes inactive, I cannot click on any of my X-windows and the skull is the mouse pointer. Finally after about 20 secs or so, the mouse pointer reverts back to the traditional pointer. But all windows had been terminated, including all my active jobs. I did a google image search on Linux Kill X Windows icon but nothing comes up matching the skull icon. How can I disable this feature so it won't be accidently invoked again? This is quite painful and I don't want to have to restart jobs which have been running for hours. Thanks. ___ From the sounds of it you may have launched the program called xkill. Your description of the dead man skull makes it sound like this is the program. It is a GUI program that essentially takes mouse input, and sends a kill command to whatever window you click on. I don't remember for sure but I think pressing ESC will cancel it. http://en.wikipedia.org/wiki/Xkill HTH Matt M. LinuxKnight ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Repairing a dead MacBook Pro
On Wed, 17 Sep 2014 14:30:18 -0700 Loren M. Lang lor...@north-winds.org wrote: I'm assuming the power supply voltages are out of spec so I pulled out my voltmeter. Measuring the SATA power connector with the dead hard drive attached and Mac powered on, I found 5V on the 5V supply lines and no voltage on the 12V or 3.3V lines. No indication of something that might kill a logic board, but I would expect to see 12V on the supply. Okay, this is wrong. There should be a +12V supply at the drive pins. It's not surprising that there isn't a +3.3V logic supply. Most drives don't use it, so most manufacturer's don't provide it. However, the 12 volt supply is required, and should always be there. Did you measure the voltages with the drive plugged in or not? The presence or absence of a load can make quite a difference. Also, did you check for an AC component on the supply pins? Or better yet, use an oscilloscope on them? It's entirely possible to have high-amplitude AC coming out of the power supply, yet still have it read the proper DC voltage with a meter. That sort of thing can fry your logic board. (Heed the Voice of Experience, my son.) Hope this helps. --Dale -- Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining. -- Jeff Raskin signature.asc Description: PGP signature ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Identifying Why One Web Site Will Not Load [RESOLVED]
On Thu, Sep 18, 2014 at 6:24 AM, Rich Shepard rshep...@appl-ecosys.com wrote: On Wed, 17 Sep 2014, Jim Garrison wrote: One possibility is that the *remote* web server has reverse DNS validation turned on, and your external IP, for some reason, does not have a PTR record. Can you send me (in a private email) your external IP address? You can easily determine this by going to http://whatismyip.com. Jim, Easily found with whois, host, and dig on the domain name. However, since you made me see what I overlooked before, the IP address for this domain is static; it points to the server hosting my web site at my ISP. It's only the mail. IP address that is dynamic since mail comes directly here. Almost all ISPs provide PTR records for their dynamically assigned pool addresses. For example, my (obfuscated) IP is aa.bb.cc.dd and if I do a dig -x on it I get: This morning a message from the GRASS coordinator said that the web server had hardware issues that slowed response times. Why my location should seem to be the only one consistently affected will probably never be known. No PTR record returned by dig, only the SOA: $ dig 216.99.193.149 ; DiG 9.9.5-P1 216.99.193.149 ;; global options: +cmd ;; Got answer: ;; -HEADER- opcode: QUERY, status: NXDOMAIN, id: 57347 ;; flags: qr rd ra ad; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 1 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 512 ;; QUESTION SECTION: ;216.99.193.149.IN A ;; AUTHORITY SECTION: . 1777IN SOA a.root-servers.net. nstld.verisign-grs.com. 2014091800 1800 900 604800 86400 ;; Query time: 38 msec ;; SERVER: 8.8.8.8#53(8.8.8.8) ;; WHEN: Thu Sep 18 06:18:42 PDT 2014 ;; MSG SIZE rcvd: 118 If your ISP has not provided a PTR record for your address, then it's the SERVER that could be timing out on DNS resolution trying to figure out who you are (for blacklisting, for instance). It's pretty rare nowadays to configure a web server to do PTR lookup (email servers yes, web servers no) but it would cause exactly the behavior you are seeing. Perhaps that's it. To my eye this dig result fits Jim's explanation perfectly. The webserver is doing a reverse DNS lookup for some reason. Maybe for blacklisting or maybe just for logging. This article describes how you can accidentally configure Apache to do this by putting in an Allow from localhost directive. http://blog.endpoint.com/2013/09/apache-accidental-dns-hostname-lookups.html This problem could resolve itself because your IP address is so dynamic. It might randomly change to an IP address that has a PTR record and then suddenly you will be able to load the website normally. Bill ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Identifying Why One Web Site Will Not Load [RESOLVED]
On Thu, 18 Sep 2014, Bill Barry wrote: To my eye this dig result fits Jim's explanation perfectly. The webserver is doing a reverse DNS lookup for some reason. Maybe for blacklisting or maybe just for logging. Bill, Yes, Jim identified a potential cause. This article describes how you can accidentally configure Apache to do this by putting in an Allow from localhost directive. http://blog.endpoint.com/2013/09/apache-accidental-dns-hostname-lookups.html I've no idea if the GRASS Web server deliberately does reverss DNS lookups or not. But, I'll pass on the information. This problem could resolve itself because your IP address is so dynamic. Not the Web server's IP address. That's at my ISP and is static. Only the MX address is dynamic. Thanks, Rich ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Identifying Why One Web Site Will Not Load [RESOLVED]
On Thu, Sep 18, 2014 at 9:35 AM, Rich Shepard rshep...@appl-ecosys.com wrote: Not the Web server's IP address. That's at my ISP and is static. Only the MX address is dynamic. The idea is that the grass.osgeo.org webserver is looking up the PTR record for the external IP address of the machine that is running your web browser. In your case it is the same as your MX address. host mail.appl-ecosys.com mail.appl-ecosys.com has address 50.38.89.212 dig -x 50.38.89.212 ; DiG 9.9.5-4-Debian -x 50.38.89.212 ;; global options: +cmd ;; Got answer: ;; -HEADER- opcode: QUERY, status: SERVFAIL, id: 48866 ;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 1 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags: do; udp: 4096 ;; QUESTION SECTION: ;212.89.38.50.in-addr.arpa. IN PTR ;; Query time: 543 msec ;; SERVER: 172.30.42.1#53(172.30.42.1) ;; WHEN: Thu Sep 18 09:39:20 PDT 2014 ;; MSG SIZE rcvd: 54 Which shows no PTR record. Bill ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Identifying Why One Web Site Will Not Load [RESOLVED]
On rereading my post, I'm not sure I made myself clear. I'm assuming you're running the browser on a local system (i.e. a computer/laptop to which you have physical access), and that system is connected to the Internet via a local network and an ISP-provided DSL/cable modem. The address I'm referring to is the routable IP of the WAN side of the modem, which is what you'll get from http://whatismyip.com If that does not describe your situation, please clarify. -- Jim Garrison (j...@acm.org) PGP Keys at http://www.jhmg.net RSA 0x04B73B7F DH 0x70738D88 ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Identifying Why One Web Site Will Not Load [RESOLVED]
On Thu, 18 Sep 2014, Bill Barry wrote: The idea is that the grass.osgeo.org webserver is looking up the PTR record for the external IP address of the machine that is running your web browser. In your case it is the same as your MX address. Bill, Ah, yes. Now I'm back on the right page. host mail.appl-ecosys.com mail.appl-ecosys.com has address 50.38.89.212 dig -x 50.38.89.212 ;212.89.38.50.in-addr.arpa. IN PTR Which shows no PTR record. Isn't the above the PTR record, or is it asking what would be the PTR record for that reverse address? Rich ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
[PLUG] Question About RAID and LVM - Ubunu 14.04
I am confused about a RAID1 and LVM installation using Ubuntu 14.04 server. I have a new System 76 Gazelle Pro laptop with two 1 TB SSD drives. I re-installed the system using the server version of 14.04 so I could have the installer create the raid and lvm. I followed this blog post - http://blog.miketoscano.com/?p=307, and the steps in the installer to create the raid and then the lvm, and then install the OS, and it all seemed to work. I then installed the ubuntu-desktop and then the system76 drivers. It all seems to be working. However, I cannot find mdadm. It does not appear to be installed. I looked at /etc/fstab and I see mark@tsunami:~$ cat /etc/fstab # /etc/fstab: static file system information. # # Use 'blkid' to print the universally unique identifier for a # device; this may be used with UUID= as a more robust way to name devices # that works even if disks are added and removed. See fstab(5). # # file system mount point type options dump pass /dev/mapper/vg1_tsunami-root / ext4 errors=remount-ro 0 1 /dev/mapper/vg1_tsunami-swap none swap sw 0 0 And the free space shows mark@tsunami:~$ df -h df: ‘/run/user/107/gvfs’: Permission denied Filesystem Size Used Avail Use% Mounted on /dev/mapper/vg1_tsunami-root 1.8T 120G 1.6T 7% / none 4.0K 0 4.0K 0% /sys/fs/cgroup udev 7.8G 4.0K 7.8G 1% /dev tmpfs 1.6G 1.2M 1.6G 1% /run none 5.0M 0 5.0M 0% /run/lock none 7.8G 144K 7.8G 1% /run/shm none 100M 40K 100M 1% /run/user There is no mdadm.conf. ]mark@tsunami:~$ sudo locate mdadm [sudo] password for mark: /usr/share/bash-completion/completions/mdadm And mdstat does not seem correct from what I have read. ]mark@tsunami:~$ cat /proc/mdstat Personalities : unused devices: none Am I missing something? Do I really have what I intended to install? I don't want to pull all my files over to the machine until I am sure it is working as intended. Thanks, Mark ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Question About RAID and LVM - Ubunu 14.04
Finally a question I might be able to help with! I'm so excited! I can think
of many ways to get more information to help alleviate the confusion. I'd
suggest starting with an LVM scan of physical volumes: pvscan -v. That
will tell you weather LVM is using 'md#' devices, or the hardware directly
('sd#' devices). You could also query the hard drives with mdadm for RAID
headers. I think the command would be mdadm --query /dev/sd#.
On September 18, 2014 1:28:23 PM PDT, Mark Phillips
m...@phillipsmarketing.biz wrote:
I am confused about a RAID1 and LVM installation using Ubuntu 14.04
server.
I have a new System 76 Gazelle Pro laptop with two 1 TB SSD drives. I
re-installed the system using the server version of 14.04 so I could
have
the installer create the raid and lvm. I followed this blog post -
http://blog.miketoscano.com/?p=307, and the steps in the installer to
create the raid and then the lvm, and then install the OS, and it all
seemed to work. I then installed the ubuntu-desktop and then the
system76
drivers. It all seems to be working.
However, I cannot find mdadm. It does not appear to be installed. I
looked
at /etc/fstab and I see
mark@tsunami:~$ cat /etc/fstab
# /etc/fstab: static file system information.
#
# Use 'blkid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust way to name
devices
# that works even if disks are added and removed. See fstab(5).
#
# file system mount point type options dump pass
/dev/mapper/vg1_tsunami-root / ext4 errors=remount-ro 0 1
/dev/mapper/vg1_tsunami-swap none swap sw 0 0
And the free space shows
mark@tsunami:~$ df -h
df: ‘/run/user/107/gvfs’: Permission denied
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/vg1_tsunami-root 1.8T 120G 1.6T 7% /
none 4.0K 0 4.0K 0% /sys/fs/cgroup
udev 7.8G 4.0K 7.8G 1% /dev
tmpfs 1.6G 1.2M 1.6G 1% /run
none 5.0M 0 5.0M 0% /run/lock
none 7.8G 144K 7.8G 1% /run/shm
none 100M 40K 100M 1% /run/user
There is no mdadm.conf.
]mark@tsunami:~$ sudo locate mdadm
[sudo] password for mark:
/usr/share/bash-completion/completions/mdadm
And mdstat does not seem correct from what I have read.
]mark@tsunami:~$ cat /proc/mdstat
Personalities :
unused devices: none
Am I missing something? Do I really have what I intended to install? I
don't want to pull all my files over to the machine until I am sure it
is
working as intended.
Thanks,
Mark
___
PLUG mailing list
PLUG@lists.pdxlinux.org
http://lists.pdxlinux.org/mailman/listinfo/plug
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.
___
PLUG mailing list
PLUG@lists.pdxlinux.org
http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Question About RAID and LVM - Ubunu 14.04
Josiah,
Thanks for your response. It looks as if LVM is using the hardware directly
-
mark@tsunami:~$ sudo pvscan -v
[sudo] password for mark:
Wiping cache of LVM-capable devices
Wiping internal VG cache
Walking through all physical volumes
PV /dev/sda1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
PV /dev/sdb1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
Total: 2 [1.82 TiB] / in use: 2 [1.82 TiB] / in no VG: 0 [0 ]
mark@tsunami:~$
Part of my confusion is that mdadm is NOT installed on my system.
Everything I have read about raid involves mdadam, but as I said in my
original post, it is not installed -
mark@tsunami:~$ sudo mdadm
sudo: mdadm: command not found
mark@tsunami:~$
However, in the Ubuntu 14.04 server installer I specifically set up the
drives to be in a raid1 array. From the instructions I followed (see
reference above)
- Designate your new partition for RAID by selecting “Physical volume
for RAID” at the “How to use this partition:” prompt. This process will
create a new RAID device.
Repeat the previous step for the other physical disk.
- Here’s the overview of my partition layout and settings: (mine looked
the same, but the drives are 1 TB Samsung drives.
-
At the prompt asking, “Write the changes to disks and configure LVM?”
Select yes.
I then entered the LVM process, finished the installation, picked a few
packages including ubuntu-desktop, and as I said above it boots just fine.
What should I do now?
I could do a re-install, but the steps won't change, so I am not confident
I will get a RAID1 array out of it.
Or, I could install mdadm and see what it says.but can a software raid
be installed without mdadm? If the system is magically configured as a raid
array, will installing mdadm screw it up?
I don't believe this laptop has any raid hardware installedat least
System76 never told me about it, and I asked them to configure the drives
as RAID1 when I bought the beast and they said they could not do that.
Thanks,
Mark
On Thu, Sep 18, 2014 at 3:45 PM, Josiah Luscher s...@josiahluscher.com
wrote:
Finally a question I might be able to help with! I'm so excited! I can
think of many ways to get more information to help alleviate the
confusion. I'd suggest starting with an LVM scan of physical volumes:
pvscan -v. That will tell you weather LVM is using 'md#' devices, or
the hardware directly ('sd#' devices). You could also query the hard
drives with mdadm for RAID headers. I think the command would be mdadm
--query /dev/sd#.
On September 18, 2014 1:28:23 PM PDT, Mark Phillips
m...@phillipsmarketing.biz wrote:
I am confused about a RAID1 and LVM installation using Ubuntu 14.04
server.
I have a new System 76 Gazelle Pro laptop with two 1 TB SSD drives. I
re-installed the system using the server version of 14.04 so I could
have
the installer create the raid and lvm. I followed this blog post -
http://blog.miketoscano.com/?p=307, and the steps in the installer to
create the raid and then the lvm, and then install the OS, and it all
seemed to work. I then installed the ubuntu-desktop and then the
system76
drivers. It all seems to be working.
However, I cannot find mdadm. It does not appear to be installed. I
looked
at /etc/fstab and I see
mark@tsunami:~$ cat /etc/fstab
# /etc/fstab: static file system information.
#
# Use 'blkid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust way to name
devices
# that works even if disks are added and removed. See fstab(5).
#
# file system mount point type options dump pass
/dev/mapper/vg1_tsunami-root / ext4 errors=remount-ro 0 1
/dev/mapper/vg1_tsunami-swap none swap sw 0 0
And the free space shows
mark@tsunami:~$ df -h
df: ‘/run/user/107/gvfs’: Permission denied
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/vg1_tsunami-root 1.8T 120G 1.6T 7% /
none 4.0K 0 4.0K 0% /sys/fs/cgroup
udev 7.8G 4.0K 7.8G 1% /dev
tmpfs 1.6G 1.2M 1.6G 1% /run
none 5.0M 0 5.0M 0% /run/lock
none 7.8G 144K 7.8G 1% /run/shm
none 100M 40K 100M 1% /run/user
There is no mdadm.conf.
]mark@tsunami:~$ sudo locate mdadm
[sudo] password for mark:
/usr/share/bash-completion/completions/mdadm
And mdstat does not seem correct from what I have read.
]mark@tsunami:~$ cat /proc/mdstat
Personalities :
unused devices: none
Am I missing something? Do I really have what I intended to install? I
don't want to pull all my files over to the machine until I am sure it
is
working as intended.
Thanks,
Mark
___
PLUG mailing list
PLUG@lists.pdxlinux.org
http://lists.pdxlinux.org/mailman/listinfo/plug
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.
___
PLUG mailing list
PLUG@lists.pdxlinux.org
http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Question About RAID and LVM - Ubunu 14.04
On Thu, 18 Sep 2014 18:17:28 -0700, Mark Phillips
m...@phillipsmarketing.biz wrote:
Josiah,
Thanks for your response. It looks as if LVM is using the hardware directly
Yes, it looks like LVM is directly using the hardware indeed. The
Ubuntu Desktop installer does not include RAID support nor install mdadm
by default. You can get mdadm by doing sudo apt-get install mdadm from
the command-line. In fact, Ubuntu would have said so if you tried
running mdadm without using sudo. It's not smart enough, though, to
recommend such things via sudo.
The real problem is that both hard drives are allocated as independent
physical volumes to LVM giving you twice the space you would normally
have. I also noticed that you have all of the volume group entirely
allocated to one physical volume. That kind of negates any benefit to
using LVM as live volumes can't be shrunk. Most benefits of LVM like
taking live snapshots or creating new volumes on the fly require that
there is some unallocated space in the volume group. There is a way to
restore RAID without reinstalling, but it's a long, complicated
procedure.
I recommend to reinstall and use the Ubuntu Server ISO. Set up RAID and
LVM as you see fit. Once Server is finished installing, just install the
ubuntu-desktop package (or kubuntu-desktop, xubuntu-desktop, etc.) and
you will have a normal desktop. Ubuntu Desktop is actually based on the
standard Server install plus everything ubuntu-desktop pulls in.
sudo apt-get install ubuntu-desktop
-
mark@tsunami:~$ sudo pvscan -v
[sudo] password for mark:
Wiping cache of LVM-capable devices
Wiping internal VG cache
Walking through all physical volumes
PV /dev/sda1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
PV /dev/sdb1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
Total: 2 [1.82 TiB] / in use: 2 [1.82 TiB] / in no VG: 0 [0 ]
mark@tsunami:~$
Part of my confusion is that mdadm is NOT installed on my system.
Everything I have read about raid involves mdadam, but as I said in my
original post, it is not installed -
mark@tsunami:~$ sudo mdadm
sudo: mdadm: command not found
mark@tsunami:~$
However, in the Ubuntu 14.04 server installer I specifically set up the
drives to be in a raid1 array. From the instructions I followed (see
reference above)
- Designate your new partition for RAID by selecting “Physical volume
for RAID” at the “How to use this partition:” prompt. This process will
create a new RAID device.
Repeat the previous step for the other physical disk.
- Here’s the overview of my partition layout and settings: (mine looked
the same, but the drives are 1 TB Samsung drives.
-
At the prompt asking, “Write the changes to disks and configure LVM?”
Select yes.
I then entered the LVM process, finished the installation, picked a few
packages including ubuntu-desktop, and as I said above it boots just fine.
What should I do now?
I could do a re-install, but the steps won't change, so I am not confident
I will get a RAID1 array out of it.
Or, I could install mdadm and see what it says.but can a software raid
be installed without mdadm? If the system is magically configured as a raid
array, will installing mdadm screw it up?
I don't believe this laptop has any raid hardware installedat least
System76 never told me about it, and I asked them to configure the drives
as RAID1 when I bought the beast and they said they could not do that.
Thanks,
Mark
On Thu, Sep 18, 2014 at 3:45 PM, Josiah Luscher s...@josiahluscher.com
wrote:
Finally a question I might be able to help with! I'm so excited! I can
think of many ways to get more information to help alleviate the
confusion. I'd suggest starting with an LVM scan of physical volumes:
pvscan -v. That will tell you weather LVM is using 'md#' devices, or
the hardware directly ('sd#' devices). You could also query the hard
drives with mdadm for RAID headers. I think the command would be mdadm
--query /dev/sd#.
On September 18, 2014 1:28:23 PM PDT, Mark Phillips
m...@phillipsmarketing.biz wrote:
I am confused about a RAID1 and LVM installation using Ubuntu 14.04
server.
I have a new System 76 Gazelle Pro laptop with two 1 TB SSD drives. I
re-installed the system using the server version of 14.04 so I could
have
the installer create the raid and lvm. I followed this blog post -
http://blog.miketoscano.com/?p=307, and the steps in the installer to
create the raid and then the lvm, and then install the OS, and it all
seemed to work. I then installed the ubuntu-desktop and then the
system76
drivers. It all seems to be working.
However, I cannot find mdadm. It does not appear to be installed. I
looked
at /etc/fstab and I see
mark@tsunami:~$ cat /etc/fstab
# /etc/fstab: static file system information.
#
# Use 'blkid' to print the universally unique identifier for a
# device; this may be used with UUID= as a more robust
Re: [PLUG] Question About RAID and LVM - Ubunu 14.04
Loren,
Thanks for you comments. As I said in my first post, I used ubuntu server
to install both raid and lvm, and then installed ubuntu-desktop. The image
in my second post is from the text based ubuntu server installation process.
I guess I could try it again, but I don't see how anything would change.
Thanks,
Mark
On Sep 18, 2014 8:54 PM, Loren M. Lang lor...@north-winds.org wrote:
On Thu, 18 Sep 2014 18:17:28 -0700, Mark Phillips
m...@phillipsmarketing.biz wrote:
Josiah,
Thanks for your response. It looks as if LVM is using the hardware
directly
Yes, it looks like LVM is directly using the hardware indeed. The
Ubuntu Desktop installer does not include RAID support nor install mdadm
by default. You can get mdadm by doing sudo apt-get install mdadm from
the command-line. In fact, Ubuntu would have said so if you tried
running mdadm without using sudo. It's not smart enough, though, to
recommend such things via sudo.
The real problem is that both hard drives are allocated as independent
physical volumes to LVM giving you twice the space you would normally
have. I also noticed that you have all of the volume group entirely
allocated to one physical volume. That kind of negates any benefit to
using LVM as live volumes can't be shrunk. Most benefits of LVM like
taking live snapshots or creating new volumes on the fly require that
there is some unallocated space in the volume group. There is a way to
restore RAID without reinstalling, but it's a long, complicated
procedure.
I recommend to reinstall and use the Ubuntu Server ISO. Set up RAID and
LVM as you see fit. Once Server is finished installing, just install the
ubuntu-desktop package (or kubuntu-desktop, xubuntu-desktop, etc.) and
you will have a normal desktop. Ubuntu Desktop is actually based on the
standard Server install plus everything ubuntu-desktop pulls in.
sudo apt-get install ubuntu-desktop
-
mark@tsunami:~$ sudo pvscan -v
[sudo] password for mark:
Wiping cache of LVM-capable devices
Wiping internal VG cache
Walking through all physical volumes
PV /dev/sda1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
PV /dev/sdb1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
Total: 2 [1.82 TiB] / in use: 2 [1.82 TiB] / in no VG: 0 [0 ]
mark@tsunami:~$
Part of my confusion is that mdadm is NOT installed on my system.
Everything I have read about raid involves mdadam, but as I said in my
original post, it is not installed -
mark@tsunami:~$ sudo mdadm
sudo: mdadm: command not found
mark@tsunami:~$
However, in the Ubuntu 14.04 server installer I specifically set up the
drives to be in a raid1 array. From the instructions I followed (see
reference above)
- Designate your new partition for RAID by selecting “Physical volume
for RAID” at the “How to use this partition:” prompt. This process
will
create a new RAID device.
Repeat the previous step for the other physical disk.
- Here’s the overview of my partition layout and settings: (mine
looked
the same, but the drives are 1 TB Samsung drives.
-
At the prompt asking, “Write the changes to disks and configure LVM?”
Select yes.
I then entered the LVM process, finished the installation, picked a few
packages including ubuntu-desktop, and as I said above it boots just
fine.
What should I do now?
I could do a re-install, but the steps won't change, so I am not
confident
I will get a RAID1 array out of it.
Or, I could install mdadm and see what it says.but can a software
raid
be installed without mdadm? If the system is magically configured as a
raid
array, will installing mdadm screw it up?
I don't believe this laptop has any raid hardware installedat least
System76 never told me about it, and I asked them to configure the drives
as RAID1 when I bought the beast and they said they could not do that.
Thanks,
Mark
On Thu, Sep 18, 2014 at 3:45 PM, Josiah Luscher s...@josiahluscher.com
wrote:
Finally a question I might be able to help with! I'm so excited! I can
think of many ways to get more information to help alleviate the
confusion. I'd suggest starting with an LVM scan of physical volumes:
pvscan -v. That will tell you weather LVM is using 'md#' devices,
or
the hardware directly ('sd#' devices). You could also query the hard
drives with mdadm for RAID headers. I think the command would be
mdadm
--query /dev/sd#.
On September 18, 2014 1:28:23 PM PDT, Mark Phillips
m...@phillipsmarketing.biz wrote:
I am confused about a RAID1 and LVM installation using Ubuntu 14.04
server.
I have a new System 76 Gazelle Pro laptop with two 1 TB SSD drives. I
re-installed the system using the server version of 14.04 so I could
have
the installer create the raid and lvm. I followed this blog post -
http://blog.miketoscano.com/?p=307, and the steps in the installer to
Re: [PLUG] Fetchmail vs. Getmail
On Sun, 31 Aug 2014 07:15:13 -0700 (PDT), Rich Shepard rshep...@appl-ecosys.com wrote: I'm looking for opinions and experience running getmail rather than fetchmail locally. My mail directory uses mbox format and my MUA is alpine. Having both in- and out-going mail relayed through my ISP might be my only option given the apparent inability of namecheap to dynamically update the IP address for the mail host here (that's the lower section of their host records definition page.) My recommendation might be to find an SMTP Redirection service. I use Rollernet and was presently surprised at how responsive their support was even when I was using a free account. I currently have upgraded to a paid Personal account. During any periods where my Internet is down or out of date with Rollernet's configuration for my IP address, they queue up the mail and I can request it to be pushed immediately once it is resolved. I can also monitor what and how much mail has queued up. As for CNAMEs, the RFCs are clear that MX, SRV, and PTR records are supposed to point to A or records. It's allowed for a CNAME to point to an MX record for mail, but the MX record should contain the name of an A (IPv4) and/or (IPv6) directly. Anything else and you are operating out of scope of the standard and stuff might break. With that said, I don't recall running into any issues, but I fixed it just in case. See RFC 5321 if you want to gory details. Rich ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug ___ PLUG mailing list PLUG@lists.pdxlinux.org http://lists.pdxlinux.org/mailman/listinfo/plug
Re: [PLUG] Question About RAID and LVM - Ubunu 14.04
If you did install raid1, then the installer didn't figure out that you
needed to include the mdadm
package. Also in raid1, either drive would show the same info if things
were correctly configured.
Loading the mdadm package and querying the drives will show you that info.
Your link about setting up raid is valid. It just doesn't say anything
about what packages should be installed.
On Thu, Sep 18, 2014 at 9:05 PM, Mark Phillips m...@phillipsmarketing.biz
wrote:
Loren,
Thanks for you comments. As I said in my first post, I used ubuntu server
to install both raid and lvm, and then installed ubuntu-desktop. The image
in my second post is from the text based ubuntu server installation
process.
I guess I could try it again, but I don't see how anything would change.
Thanks,
Mark
On Sep 18, 2014 8:54 PM, Loren M. Lang lor...@north-winds.org wrote:
On Thu, 18 Sep 2014 18:17:28 -0700, Mark Phillips
m...@phillipsmarketing.biz wrote:
Josiah,
Thanks for your response. It looks as if LVM is using the hardware
directly
Yes, it looks like LVM is directly using the hardware indeed. The
Ubuntu Desktop installer does not include RAID support nor install mdadm
by default. You can get mdadm by doing sudo apt-get install mdadm from
the command-line. In fact, Ubuntu would have said so if you tried
running mdadm without using sudo. It's not smart enough, though, to
recommend such things via sudo.
The real problem is that both hard drives are allocated as independent
physical volumes to LVM giving you twice the space you would normally
have. I also noticed that you have all of the volume group entirely
allocated to one physical volume. That kind of negates any benefit to
using LVM as live volumes can't be shrunk. Most benefits of LVM like
taking live snapshots or creating new volumes on the fly require that
there is some unallocated space in the volume group. There is a way to
restore RAID without reinstalling, but it's a long, complicated
procedure.
I recommend to reinstall and use the Ubuntu Server ISO. Set up RAID and
LVM as you see fit. Once Server is finished installing, just install the
ubuntu-desktop package (or kubuntu-desktop, xubuntu-desktop, etc.) and
you will have a normal desktop. Ubuntu Desktop is actually based on the
standard Server install plus everything ubuntu-desktop pulls in.
sudo apt-get install ubuntu-desktop
-
mark@tsunami:~$ sudo pvscan -v
[sudo] password for mark:
Wiping cache of LVM-capable devices
Wiping internal VG cache
Walking through all physical volumes
PV /dev/sda1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
PV /dev/sdb1 VG vg1_tsunami lvm2 [931.51 GiB / 0free]
Total: 2 [1.82 TiB] / in use: 2 [1.82 TiB] / in no VG: 0 [0 ]
mark@tsunami:~$
Part of my confusion is that mdadm is NOT installed on my system.
Everything I have read about raid involves mdadam, but as I said in my
original post, it is not installed -
mark@tsunami:~$ sudo mdadm
sudo: mdadm: command not found
mark@tsunami:~$
However, in the Ubuntu 14.04 server installer I specifically set up the
drives to be in a raid1 array. From the instructions I followed (see
reference above)
- Designate your new partition for RAID by selecting “Physical
volume
for RAID” at the “How to use this partition:” prompt. This process
will
create a new RAID device.
Repeat the previous step for the other physical disk.
- Here’s the overview of my partition layout and settings: (mine
looked
the same, but the drives are 1 TB Samsung drives.
-
At the prompt asking, “Write the changes to disks and configure
LVM?”
Select yes.
I then entered the LVM process, finished the installation, picked a few
packages including ubuntu-desktop, and as I said above it boots just
fine.
What should I do now?
I could do a re-install, but the steps won't change, so I am not
confident
I will get a RAID1 array out of it.
Or, I could install mdadm and see what it says.but can a software
raid
be installed without mdadm? If the system is magically configured as a
raid
array, will installing mdadm screw it up?
I don't believe this laptop has any raid hardware installedat least
System76 never told me about it, and I asked them to configure the
drives
as RAID1 when I bought the beast and they said they could not do that.
Thanks,
Mark
On Thu, Sep 18, 2014 at 3:45 PM, Josiah Luscher s...@josiahluscher.com
wrote:
Finally a question I might be able to help with! I'm so excited! I
can
think of many ways to get more information to help alleviate the
confusion. I'd suggest starting with an LVM scan of physical volumes:
pvscan -v. That will tell you weather LVM is using 'md#'
devices,
or
the hardware directly ('sd#' devices). You
