[opensuse] Any problems with installing 12.1 on RAID ?
I've just started a remote install of 12.1 on a server with pre-configured RAID and LVM. In YaST, md0 and md1 for some reason turned up under "harddisk", whereas md2 is listed under RAID. I can't select md0 for swap nor md1 for root. I've tried a rescan, but that didn't help. The logival volumes are fine. Any known problems with RAID and 12.1? Unfortunately I don't have the time to investigate, this server needs to be up and running, so I guess 11.4 + zypper dup is the next best option. -- Per Jessen, Zürich (6.5°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Per Jessen wrote:
I've just started a remote install of 12.1 on a server with pre-configured RAID and LVM. In YaST, md0 and md1 for some reason turned up under "harddisk", whereas md2 is listed under RAID. I can't select md0 for swap nor md1 for root. I've tried a rescan, but that didn't help. The logival volumes are fine. Any known problems with RAID and 12.1? Unfortunately I don't have the time to investigate, this server needs to be up and running, so I guess 11.4 + zypper dup is the next best option.
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott wrote:
Per Jessen wrote:
I've just started a remote install of 12.1 on a server with pre-configured RAID and LVM. In YaST, md0 and md1 for some reason turned up under "harddisk", whereas md2 is listed under RAID. I can't select md0 for swap nor md1 for root. I've tried a rescan, but that didn't help. The logival volumes are fine. Any known problems with RAID and 12.1? Unfortunately I don't have the time to investigate, this server needs to be up and running, so I guess 11.4 + zypper dup is the next best option.
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID.
My intentions were to have swap on md0, root on md1 and miscellaneous on lvs on md2. Hmm, weird. I'll capture y2logs and then proceed with 11.4 I think. -- Per Jessen, Zürich (7.2°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Per Jessen wrote:
James Knott wrote:
Per Jessen wrote:
I've just started a remote install of 12.1 on a server with pre-configured RAID and LVM. In YaST, md0 and md1 for some reason turned up under "harddisk", whereas md2 is listed under RAID. I can't select md0 for swap nor md1 for root. I've tried a rescan, but that didn't help. The logival volumes are fine. Any known problems with RAID and 12.1? Unfortunately I don't have the time to investigate, this server needs to be up and running, so I guess 11.4 + zypper dup is the next best option.
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID.
My intentions were to have swap on md0, root on md1 and miscellaneous on lvs on md2. Hmm, weird. I'll capture y2logs and then proceed with 11.4 I think.
Weirder and weirder - 11.4 shows roughly the same. I guess I haven't done any recent installations on to software RAID. Is there a way of manually editing a file to set up formatting, mount point etc? -- Per Jessen, Zürich (7.4°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Per Jessen wrote:
Per Jessen wrote:
James Knott wrote:
Per Jessen wrote:
I've just started a remote install of 12.1 on a server with pre-configured RAID and LVM. In YaST, md0 and md1 for some reason turned up under "harddisk", whereas md2 is listed under RAID. I can't select md0 for swap nor md1 for root. I've tried a rescan, but that didn't help. The logival volumes are fine. Any known problems with RAID and 12.1? Unfortunately I don't have the time to investigate, this server needs to be up and running, so I guess 11.4 + zypper dup is the next best option.
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID.
My intentions were to have swap on md0, root on md1 and miscellaneous on lvs on md2. Hmm, weird. I'll capture y2logs and then proceed with 11.4 I think.
Weirder and weirder - 11.4 shows roughly the same. I guess I haven't done any recent installations on to software RAID. Is there a way of manually editing a file to set up formatting, mount point etc?
I reported this issue a little less than a year ago: https://bugzilla.novell.com/show_bug.cgi?id=660427 -- Per Jessen, Zürich (7.6°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott said the following on 12/15/2011 11:09 AM:
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID.
Perhaps you should write that up for the HOW-TO/Wiki That's useful stuff. -- The eye sees only what the mind is prepared to comprehend. -- Henry Bergson, French Philosopher (1859-1941). -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Anton Aylward wrote:
James Knott said the following on 12/15/2011 11:09 AM:
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID.
Perhaps you should write that up for the HOW-TO/Wiki
That's useful stuff.
There's really nothing much to it, I've been running systems like that for 5+ years. Boot from RAID1, root on RAID1. -- Per Jessen, Zürich (6.8°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Per Jessen said the following on 12/15/2011 12:04 PM:
Anton Aylward wrote:
James Knott said the following on 12/15/2011 11:09 AM:
Last week, I installed it on a RAID1 & LVM array. No problems. Even /boot is on RAID.
Perhaps you should write that up for the HOW-TO/Wiki
That's useful stuff.
There's really nothing much to it, I've been running systems like that for 5+ years. Boot from RAID1, root on RAID1.
I was thinking of the context "SystemD under OpenSuse 12.1" Things that worked well for years (5+ possibly) in SysVInit systems have to be reconsidered under SystemD. We've a number of threads here about that :-) -- For every complex problem there is an answer that is clear, simple, and wrong. --Henry Louis Mencken -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Anton Aylward wrote:
I was thinking of the context "SystemD under OpenSuse 12.1"
As I've mentioned, I find systemd too flakey to use. Of the 3 computers I've installed 12.1 on, 2 of them have problems getting networking to run in systemd. The 3rd, is a notebook that uses network management plasmoid, instead of ifup, for networking. Get networking working reliably and also fetchmail to work at all in systemd, then we can talk about running a server etc. on it. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott said the following on 12/15/2011 12:29 PM:
Anton Aylward wrote:
I was thinking of the context "SystemD under OpenSuse 12.1"
As I've mentioned, I find systemd too flakey to use.
What you mean is I (James Knott) find the implementation of SystemD on openSuse 12.1 too flakey to use. I'll grant you that. But I (Anton Aylward) find the implementation of SystemD on Fedora 15 rock solid and easy to use. So I don't think the problem is with SystemD per-se but rather the implementation of it on openSuse 12.1 Damnittohell, this is the same kind of "alpha" stuff we had when KDE4.1 came out.
Of the 3 computers I've installed 12.1 on, 2 of them have problems getting networking to run in systemd. The 3rd, is a notebook that uses network management plasmoid, instead of ifup, for networking. Get networking working reliably and also fetchmail to work at all in systemd, then we can talk about running a server etc. on it.
How many of those are production machines and how many are 'scratch' machines? I hope none are critical for 'production'. One 'but' I saw reported for Fedora _before_ I installed it, and presumably it was fixed by the time I did install, was that the BASH shell was missing from /etc/shells. This will have/would have cascading consequences. Among those are that the "compatibility" with the scripts under /etc/rc.d gets broken. Personally I think that having this 'backwards compatibility' was a mistake. SystemD should not have been released yet, not until it was a complete change-over, not until everything is running under SystemD services and not a kludge that sometimes runs the old scripts 'cos no-one has got around to completing the stuff that SystemD needs to start up Postfix/Sendmail etc etc. -- Teachers open the door. You enter by yourself. Chinese Proverb -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Anton Aylward wrote:
But
I (Anton Aylward) find the implementation of SystemD on Fedora 15 rock solid and easy to use.
Do you use ifup for networking? Run fetchmail?
So I don't think the problem is with SystemD per-se but rather the implementation of it on openSuse 12.1
Damnittohell, this is the same kind of "alpha" stuff we had when KDE4.1 came out.
There was also the problem with network manager plasmoid that requires root password to create any WiFi connections. This wasn't a software error. It was an incredibly stupid design decision. I also have one computer that I have to boot through the install disc.
Of the 3 computers I've installed 12.1 on, 2 of them have problems getting networking to run in systemd. The 3rd, is a notebook that uses network management plasmoid, instead of ifup, for networking. Get networking working reliably and also fetchmail to work at all in systemd, then we can talk about running a server etc. on it. How many of those are production machines and how many are 'scratch' machines? I hope none are critical for 'production'.
One is my main desktop, one my new notebook and only one is a "test" system, where I experiment with server configuration.
Personally I think that having this 'backwards compatibility' was a mistake. SystemD should not have been released yet, not until it was a complete change-over, not until everything is running under SystemD services and not a kludge that sometimes runs the old scripts 'cos no-one has got around to completing the stuff that SystemD needs to start up Postfix/Sendmail etc etc.
I find it surprising that postfix wasn't working "out of the box". -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott wrote:
One is my main desktop, one my new notebook and only one is a "test" system, where I experiment with server configuration.
I should have mentioned that the notebook uses network management plasmoid and doesn't appear to have issues with systemd, other than starting postfix. My other systems run fine if I boot into system v, other that the one requiring the install disc to boot. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott said the following on 12/15/2011 04:30 PM:
Do you use ifup for networking?
Damned if I know; I just installed Fedora-15 and it "Just Worked"
Run fetchmail?
If you read my postings here you'd know the answer ...
There was also the problem with network manager plasmoid that requires root password to create any WiFi connections. This wasn't a software error. It was an incredibly stupid design decision.
Indeed. There's a lot of that about!
I also have one computer that I have to boot through the install disc.
All that stuff about where the pre /boot boot stuff lives, MBR, which partition and how it all links together confuses me. My laptop will boot from a USB but only if you reconfigure the BIOS while the USB is there. You can't tell it try USB, then if that's not there or fails try CD/DVD, then if that's not there or fails try the hard drive regardless. And don't ask me how things get from grub to a working system under SystemD now I can't trace the scripts starting form the entry in /etc/inittab. And no, the docco isn't up t date on that.
Of the 3 computers I've installed 12.1 on, 2 of them have problems getting networking to run in systemd. The 3rd, is a notebook that uses network management plasmoid, instead of ifup, for networking. Get networking working reliably and also fetchmail to work at all in systemd, then we can talk about running a server etc. on it.
We definitely need a better logging of the boot and startup with SystemD. Try editing /etc/systemd/system.conf to turn it on :-)
I find it surprising that postfix wasn't working "out of the box".
I do too. But then I have a separate mail hub so mail capability on my workstation or laptop is a non issue. Design Decisions again. -- "Now look," Forrester said patiently, "progress is an outmoded idea. We've got to be in step with the times. We've got to ask ourselves what progress ever did for us." -- Randall Garett, "Pagan Passions", 1959 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-12-15 at 17:41 -0500, Anton Aylward wrote:
We definitely need a better logging of the boot and startup with SystemD.
It goes to /var/log/messages, intentionally not to /var/log/boot.msg
Try editing /etc/systemd/system.conf to turn it on :-)
What do you mean? - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk7qhlQACgkQtTMYHG2NR9WIGgCfWu8QzZD9EQc9ygVBandcRkpa gLQAn29XZlXkw+DzYuOh5XSW2e5T4WSR =KPNj -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carlos E. R. said the following on 12/15/2011 06:44 PM:
On Thursday, 2011-12-15 at 17:41 -0500, Anton Aylward wrote:
We definitely need a better logging of the boot and startup with SystemD.
It goes to /var/log/messages, intentionally not to /var/log/boot.msg
Ah. Help me here. I seem to recall that the internal message buffer DURING boot gets written out somewhere. My concern is that the /var file system may bot be mounted during boot :-/
Try editing /etc/systemd/system.conf to turn it on :-)
What do you mean?
Well, like what are the entries you have at the moment? Is there something like LogLevel=info in there? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-12-15 at 19:07 -0500, Anton Aylward wrote:
Carlos E. R. said the following on 12/15/2011 06:44 PM:
It goes to /var/log/messages, intentionally not to /var/log/boot.msg
(I know because I reported a bug on that)
Ah. Help me here. I seem to recall that the internal message buffer DURING boot gets written out somewhere.
Yes, with traditional systemv it was dumped to /var/log/boot.msg somehow, from memory.
My concern is that the /var file system may bot be mounted during boot :-/
True, and root is mounted read only at start, so that fsck can be done on it. But they do it, somehow. Systemd is a mystery to me. And policy kit, and udev and...
Try editing /etc/systemd/system.conf to turn it on :-)
What do you mean?
Well, like what are the entries you have at the moment? Is there something like
Default.
LogLevel=info
in there?
Let me see... [Manager] #LogLevel=info #LogTarget=syslog-or-kmsg #LogColor=yes #LogLocation=no #DumpCore=yes #CrashShell=no #ShowStatus=yes #SysVConsole=yes #CrashChVT=1 #CPUAffinity=1 2 #MountAuto=yes #SwapAuto=yes #DefaultControllers=cpu #DefaultStandardOutput=syslog #DefaultStandardError=inherit #JoinControllers=cpu,cpuacct All commented out, and no documentation comments, as other classical configuration files . What does that "LogLevel=info" do when activated? When I report a bug and they want of me a verbose log, they ask me to boot with these parameters: systemd.log_level=debug systemd.log_target=kmsg For me, that is such a nuisance that I created a separate entry on menu.lst with those parameters. It is a nuisance typing that at boot because my keyboard is not US, so that '=' and '_' change location and I have to find them by trial an error. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk7qj4MACgkQtTMYHG2NR9WengCcCbpUBSp9GObRJOqWCZU5R6Nw oIUAn0zK8/8wx40tPhmbP9CdI9zkqUBp =9jgQ -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carlos E. R. said the following on 12/15/2011 07:23 PM:
All commented out, and no documentation comments, as other classical configuration files .
Yup! And guess what! The docco for system.conf is in systemD.conf One would think that those commented values are the defaults but that "syslog-or-kmsg" would seem to indicate not ... Anyway, try uncommenting and changing the 'info' to ''debug'
What does that "LogLevel=info" do when activated?
When I report a bug and they want of me a verbose log, they ask me to boot with these parameters:
systemd.log_level=debug systemd.log_target=kmsg
Well at least that answers the "debug" question :-) -- Cats regard people as warm-blooded furniture. --Jacquelyn Mitchard -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-12-15 at 19:44 -0500, Anton Aylward wrote:
Carlos E. R. said the following on 12/15/2011 07:23 PM:
All commented out, and no documentation comments, as other classical configuration files .
Yup! And guess what! The docco for system.conf is in systemD.conf
I don't have that file, unless it is "man systemd.conf"
One would think that those commented values are the defaults but that "syslog-or-kmsg" would seem to indicate not ...
Anyway, try uncommenting and changing the 'info' to ''debug'
No visual difference, unless it is in the logs.
Well at least that answers the "debug" question :-)
Yep. Not easy to remember tokens. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk7ql4kACgkQtTMYHG2NR9XEvwCdEqegPVhJ2FXTJwT5T92QzEuI P/kAn1agUbMBeNmwLwWLGKLHCUCl9BEx =30t/ -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-12-15 at 14:13 -0500, Anton Aylward wrote:
How many of those are production machines and how many are 'scratch' machines? I hope none are critical for 'production'.
I have only one test machine with 12.1, and is a virtual one. Perhaps a dozen bugs reported, some because of systemd and others because of gnome 3. Quite a lot for a single install. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk7qfEYACgkQtTMYHG2NR9UnUQCgiWGOQXr8ZSFgVVveO9nRmKfN N7IAmwYIby/ePUUxfsksMnmVLh1uTt4D =Ai1S -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Anton Aylward wrote:
James Knott said the following on 12/15/2011 11:09 AM:
Last week, I installed it on a RAID1& LVM array. No problems. Even /boot is on RAID. Perhaps you should write that up for the HOW-TO/Wiki
That's useful stuff.
There's not much to write up, it just worked. One thing to remember is that /boot cannot be on RAID5. It has to be on RAID1 or no RAID. In the past, I've giving /boot it's own partition and used RAID5 for the rest. However, I thought I'd try it on RAID1 as an experiment. Next thing to try is enabling it to boot with one drive missing (failed). I'm experimenting on an old IBM Netfinity server, with 6 SCSI drive slots. I've got 4 - 17 GB drives, 2 68 GB and 2 137 GB, so I have lots to play with. ;-) The current configuration is everything but /home on the RAID1 & LVM array on the 2 68 GB drives and /home on a single 137 GB drive. The only problem with this server is it's a bit slow. It has dual PIII CPUs and 1 GB memory. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott said the following on 12/15/2011 12:11 PM:
There's not much to write up, it just worked.
That's been my experience with SystemD on Fedora-15, so I get to wonder why people have problems doing on 12.1 what works with no trouble here. Surely its the same code base? All I can think of is that there are issues with integration that are lacking, but ... what? -- A good landing is one you can walk away from. A great landing is one from which the plane can take off again. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Anton Aylward wrote:
James Knott said the following on 12/15/2011 12:11 PM:
There's not much to write up, it just worked.
That's been my experience with SystemD on Fedora-15, so I get to wonder why people have problems doing on 12.1 what works with no trouble here.
Surely its the same code base? All I can think of is that there are issues with integration that are lacking, but ... what?
Lack of testing? The bug about postfix not being started is so basic, it should have been caught in a test. -- Per Jessen, Zürich (7.3°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-12-15 at 18:28 +0100, Per Jessen wrote:
Lack of testing? The bug about postfix not being started is so basic, it should have been caught in a test.
When I test factory, I do not test postfix, because it is complex and I want my mail in my work setup, not on a test one that could corrupt my email. Instead I setup Thunderbird on imap if I want to read or post something. I would not have catchd the problem with postfix. However, my virtual 12.1 test machine does have postfix running. No idea if it works. [...] No, it doesn't. Elanor:~ # rcpostfix status redirecting to systemctl postfix.service - Postfix Mail Transport Agent Loaded: loaded (/lib/systemd/system/postfix.service; enabled) Active: inactive (dead) CGroup: name=systemd:/system/postfix.service Elanor:~ # chkconfig postfix Note: Forwarding request to 'systemctl is-enabled postfix.service'. enabled postfix on It could be off because in my sustem the network reports to have failed, even though it works. Dunno. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk7qgpAACgkQtTMYHG2NR9UKdACfVUA5h8WWcTtq8B1JbgIFnYba 2OEAn34X6hlaRlWIPpDVrQU/fgtQDkft =CDaa -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carlos E. R. wrote:
On Thursday, 2011-12-15 at 18:28 +0100, Per Jessen wrote:
Lack of testing? The bug about postfix not being started is so basic, it should have been caught in a test.
When I test factory, I do not test postfix, because it is complex and
Personally I don't test postfix because it is such a basic component that it just _has_ to work. Unfortunately, we have had several regressions this time: Bugs 730787, 737260, 721713.
I would not have catchd the problem with postfix.
Nor did I, not until I began wondering why certain notifications had stopped, and I started looking in the log files. -- Per Jessen, Zürich (8.9°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Anton Aylward wrote:
That's been my experience with SystemD on Fedora-15, so I get to wonder why people have problems doing on 12.1 what works with no trouble here.
I think it depends a lot on what they're doing. It works OK on my ThinkPad, where I use network manager plasmoid and not fetchmail. The two systems that use ifup both fail to start networking. Most people wouldn't be using fetchmail either. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Thursday 15 December 2011, Anton Aylward wrote:
James Knott said the following on 12/15/2011 12:11 PM:
There's not much to write up, it just worked.
That's been my experience with SystemD on Fedora-15, so I get to wonder why people have problems doing on 12.1 what works with no trouble here.
Surely its the same code base? All I can think of is that there are issues with integration that are lacking, but ... what?
I guess the reason is because systemd developers are RedHat/Fedora guys. So upstream has developed and tested on Fedora only while they luckily don't even need documentation to do things right. cu, Rudi -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Rüdiger Meier said the following on 12/15/2011 06:33 PM:
I guess the reason is because systemd developers are RedHat/Fedora guys. So upstream has developed and tested on Fedora only while they luckily don't even need documentation to do things right.
They have produced a LOT of documentation ... ... *IF* you are willing to google and are willing to read wikis and blogs and things like that ... -- Wherever you see a successful business, someone once made a courageous decision. --Peter F. Drucker -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Thu, Dec 15, 2011 at 6:33 PM, Rüdiger Meier
On Thursday 15 December 2011, Anton Aylward wrote:
James Knott said the following on 12/15/2011 12:11 PM:
There's not much to write up, it just worked.
That's been my experience with SystemD on Fedora-15, so I get to wonder why people have problems doing on 12.1 what works with no trouble here.
Surely its the same code base? All I can think of is that there are issues with integration that are lacking, but ... what?
I guess the reason is because systemd developers are RedHat/Fedora guys. So upstream has developed and tested on Fedora only while they luckily don't even need documentation to do things right.
cu, Rudi
It's my understanding a lot of upstream packages don't come with boot/init scripts. So distro's like openSUSE over the years have created their own in a lot of cases. And with systemd work is required to get that all straightened out again. My impression is that its relatively easy to get systemd to run the scripts from /etc/init.d, but the integration still has to be done one at a time. And that integration is typically not shared by the distros via upstream packages. For opensuse it seems lots of that is only now getting handled. Longer term, opensuse packagers will start writing systemd native init logic. When that happens, we will start to see bugs because new logic will get added to the systemd script and not to the /etc/init.d scripts. I wouldn't be surprised if by 12.3 or so systemd is the more reliable of the 2 init solutions. But I admit I'm going to run sysvinit with 11.1. Greg -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, 2011-12-15 at 19:40 -0500, Greg Freemyer wrote:
It's my understanding a lot of upstream packages don't come with boot/init scripts. So distro's like openSUSE over the years have created their own in a lot of cases.
And with systemd work is required to get that all straightened out again.
It is probably worse for proprietary packages, they are slow to react to changes. I'm thinking of vmware, for example. But also things most of us did, like using boot.local or after.local, get too complicated. - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk7qmboACgkQtTMYHG2NR9X9kwCeOaGzoELxq3wDUKosi9vjIpcm je8An3zTCVwQlsU5ri+oUr/+BCANxq8L =Nn9a -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott wrote:
Anton Aylward wrote:
James Knott said the following on 12/15/2011 11:09 AM:
Last week, I installed it on a RAID1& LVM array. No problems. Even /boot is on RAID. Perhaps you should write that up for the HOW-TO/Wiki
That's useful stuff.
There's not much to write up, it just worked. One thing to remember is that /boot cannot be on RAID5. It has to be on RAID1 or no RAID. In the past, I've giving /boot it's own partition and used RAID5 for the rest. However, I thought I'd try it on RAID1 as an experiment. Next thing to try is enabling it to boot with one drive missing (failed).
I'm experimenting on an old IBM Netfinity server, with 6 SCSI drive slots. I've got 4 - 17 GB drives, 2 68 GB and 2 137 GB, so I have lots to play with. ;-)
The current configuration is everything but /home on the RAID1 & LVM array on the 2 68 GB drives and /home on a single 137 GB drive.
Is any of this software RAID or are you using a (presumably) builtin hardware RAID array?
The only problem with this server is it's a bit slow. It has dual PIII CPUs and 1 GB memory.
Heh, my mail test-cluster is still running a set of PII 400-MHz with 384Mb memory each. More than plenty :-) -- Per Jessen, Zürich (7.4°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Per Jessen wrote:
Is any of this software RAID or are you using a (presumably) builtin hardware RAID array?
The only problem with this server is it's a bit slow. It has dual PIII CPUs and 1 GB memory. Heh, my mail test-cluster is still running a set of PII 400-MHz with 384Mb memory each. More than plenty:-) I ran 11.0 on this box previously. 12.1 is much slower on it. Currently my fastest system is my ThinkPad E520, which has dual Intel i3-2310M CPUs & hyperthreading and 4 GB of memory. Next up is an older 64 bit AMD CPU & 4 GB system, which is my main desktop computer. On
It's software RAID. Hardware RAID requires an optional controller. IIRC, /boot on RAID5 is fine, provided you have hardware RAID. this system 12.1 is slow enough that I'm seriously considering getting a new mom board. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott wrote:
Per Jessen wrote:
Is any of this software RAID or are you using a (presumably) builtin hardware RAID array?
It's software RAID. Hardware RAID requires an optional controller. IIRC, /boot on RAID5 is fine, provided you have hardware RAID.
I thought perhaps your Netfinity box had one by default, my Compaq/HP boxes have always had one. Check out https://bugzilla.novell.com/show_bug.cgi?id=660427 - I don't understand why you have not hit that one. (unless you've unknowingly used the work-around). -- Per Jessen, Zürich (7.3°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Per Jessen wrote:
Check outhttps://bugzilla.novell.com/show_bug.cgi?id=660427 - I don't understand why you have not hit that one. (unless you've unknowingly used the work-around).
I didn't put a file system on RAID. I put LVM on and then created the partitions. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
James Knott wrote:
Per Jessen wrote:
Check outhttps://bugzilla.novell.com/show_bug.cgi?id=660427 - I don't understand why you have not hit that one. (unless you've unknowingly used the work-around).
I didn't put a file system on RAID. I put LVM on and then created the partitions.
Ah, that's enough - that is also why my md2 was fine, but my md0 and md1 weren't. -- Per Jessen, Zürich (6.3°C) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (6)
-
Anton Aylward
-
Carlos E. R.
-
Greg Freemyer
-
James Knott
-
Per Jessen
-
Rüdiger Meier