IN OpenSuse 15.2 I need to umount the / partition but I am getting device busy errors which block umount. Its Ext4 so I should be able to fsck this problem once unmounted. Then I can finish backing up the /home and proceed to upgrade to 15.5. How do I correct this device busy problem? The system crashed and shut off during a Firefox secession I think I had too many tabs open. CWSIV
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
IN OpenSuse 15.2 I need to umount the / partition but I am getting device busy errors which block umount.
Its Ext4 so I should be able to fsck this problem once unmounted. Then I can finish backing up the /home and proceed to upgrade to 15.5.
How do I correct this device busy problem?
The system crashed and shut off during a Firefox secession I think I had too many tabs open.
CWSIV
Hmmm, When you say "crashed", were you able to log out from your user-session, or did it take the system down? (that really shouldn't happen) The "device busy" means the system believes there are file open on / that prevent an unmount. (I"m not sure you can unmount / after boot except if you are dropped to a maintenance shell) If you have a login prompt, I'd login directly as root (or drop to maintenance mode). The root user has a 5% holdback on disk space that if FF we nuts and filled the drive in its crash - root may be able to login where a user couldn't. Barring that, I'd just boot from an install disk and use the recovery tools. fsck should be present on any of the install disks. After fsck, you can chroot the system and see if it thinks it is happy again before actually trying to boot it. (at least you can check for any errors, etc.) Any other error messages in any of the logs other than the "device busy" indication you get attempting to unmount /? (which again I don't think can do to a running system outside of booting from another disk or being in maintenance mode -- others will have to fill in here) -- David C. Rankin, J.D.,P.E.
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
IN OpenSuse 15.2 I need to umount the / partition but I am getting device busy errors which block umount.
Its Ext4 so I should be able to fsck this problem once unmounted. Then I can finish backing up the /home and proceed to upgrade to 15.5.
How do I correct this device busy problem?
The system crashed and shut off during a Firefox secession I think I had too many tabs open.
CWSIV
Hmmm,
When you say "crashed", were you able to log out from your user-session, or did it take the system down? (that really shouldn't happen)
Thought it was the surge protector going out. But that would not affect booting up and the other system is on it and ok. Not thinking straight lost two cats to disease one cancer the other FIV. Also loosing my 15.2 install disk is also not a good idea. Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open. I could not find my install disk but did find the iso online so Ill go that way. The main thing is doing backup even with Knoppix to an external backup then a , long overdue, clean install. Thankfully I have another system old 32 bit but I can download and burn dvd from it. CWSIV
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss. <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso> -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On Sun, 2023-07-23 at 12:20 +0200, Carlos E. R. wrote:
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss.
First get it working and backup lest I have to pull out the drive and do it the hard way. Ill save that link as from the online it seems 15.4 is going down so Ill do 15.5 after the backup is done. CWSIV
On 2023-07-24 03:09, Carl Spitzer {L Juno} wrote:
On Sun, 2023-07-23 at 12:20 +0200, Carlos E. R. wrote:
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss.
<http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-... <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso>>
First get it working and backup
To get it working you need to run fsck from the rescue media on the link above. It is a powerful rescue image, not install media. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On Mon, 2023-07-24 at 12:40 +0200, Carlos E. R. wrote:
On 2023-07-24 03:09, Carl Spitzer {L Juno} wrote:
On Sun, 2023-07-23 at 12:20 +0200, Carlos E. R. wrote:
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss.
<http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-... <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso>>
First get it working and backup
To get it working you need to run fsck from the rescue media on the link above.
It is a powerful rescue image, not install media.
Does it work on a 15.2 system? Are their instructions on the disk because the 15.2 rescue / installdisk had none? Funny thing the KNOPPIX 9.1 and 8,6,1 did not come up in graphical mode allowing me to mount. Not sure if they can handle NVIDIA on the Dell N5150? CWSIV
On 2023-07-24 14:58, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:40 +0200, Carlos E. R. wrote:
On 2023-07-24 03:09, Carl Spitzer {L Juno} wrote:
On Sun, 2023-07-23 at 12:20 +0200, Carlos E. R. wrote:
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss.
<http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-... <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso> <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-... <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso>>>
First get it working and backup
To get it working you need to run fsck from the rescue media on the link above.
It is a powerful rescue image, not install media.
Does it work on a 15.2 system?
Certainly. You are only doing an fsck. But if that worries you, you can use the 15.2 image instead: <http://download.opensuse.org/distribution/leap/15.2/live/openSUSE-Leap-15.2-Rescue-CD-x86_64-Build31.632-Media.iso>
Are their instructions on the disk because the 15.2 rescue / installdisk had none?
Instructions? It is just a Linux live, complete, graphical mode, xfce. You are going to do "fsck /dev/sdXY" in a terminal. You have "man fsck". You have internet, mail, browser...
Funny thing the KNOPPIX 9.1 and 8,6,1 did not come up in graphical mode allowing me to mount. Not sure if they can handle NVIDIA on the Dell N5150?
That I do not know. Better chances with 15.5 than 15.2, I guess. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On Mon, 2023-07-24 at 20:35 +0200, Carlos E. R. wrote:
On 2023-07-24 14:58, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:40 +0200, Carlos E. R. wrote:
On 2023-07-24 03:09, Carl Spitzer {L Juno} wrote:
On Sun, 2023-07-23 at 12:20 +0200, Carlos E. R. wrote:
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote: > On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss.
<http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-... <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso> <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-... <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso>>>
First get it working and backup
To get it working you need to run fsck from the rescue media on the link above.
It is a powerful rescue image, not install media.
Does it work on a 15.2 system?
Certainly. You are only doing an fsck. But if that worries you, you can use the 15.2 image instead:
Are their instructions on the disk because the 15.2 rescue / installdisk had none?
Instructions?
It is just a Linux live, complete, graphical mode, xfce. You are going to do "fsck /dev/sdXY" in a terminal. You have "man fsck". You have internet, mail, browser...
The problem is graphical mode does not come up in Knoppix. I am beginning to suspect its the graphics chip. The final test will be a 15.5 install attempt after I finish backing up to external EXT4 based drive. I can not get the konsole mode of knoppix to recognize the hd on the laptop. CWSIV
On 7/24/23 18:01, Carl Spitzer {L Juno} wrote:
The problem is graphical mode does not come up in Knoppix. I am beginning to suspect its the graphics chip. The final test will be a 15.5 install attempt after I finish backing up to external EXT4 based drive.
I can not get the konsole mode of knoppix to recognize the hd on the laptop.
That's odd. Did you run "fdisk -l" and hd didn't show up? Regards, Lew
On 2023-07-25 03:08, Lew Wolfgang wrote:
On 7/24/23 18:01, Carl Spitzer {L Juno} wrote:
The problem is graphical mode does not come up in Knoppix. I am beginning to suspect its the graphics chip. The final test will be a 15.5 install attempt after I finish backing up to external EXT4 based drive.
I can not get the konsole mode of knoppix to recognize the hd on the laptop.
That's odd. Did you run "fdisk -l" and hd didn't show up?
Mounting fails if the partition needs fsck. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On 7/25/23 03:35, Carlos E. R. wrote:
On 2023-07-25 03:08, Lew Wolfgang wrote:
On 7/24/23 18:01, Carl Spitzer {L Juno} wrote:
The problem is graphical mode does not come up in Knoppix. I am beginning to suspect its the graphics chip. The final test will be a 15.5 install attempt after I finish backing up to external EXT4 based drive.
I can not get the konsole mode of knoppix to recognize the hd on the laptop.
That's odd. Did you run "fdisk -l" and hd didn't show up?
Mounting fails if the partition needs fsck.
My bad! I meant to type fdisk -l, run from a rescue system. If fdisk can't see the disk something's really wrong. Regards, Lew
On 2023-07-25 03:01, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 20:35 +0200, Carlos E. R. wrote:
On 2023-07-24 14:58, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:40 +0200, Carlos E. R. wrote:
On 2023-07-24 03:09, Carl Spitzer {L Juno} wrote:
On Sun, 2023-07-23 at 12:20 +0200, Carlos E. R. wrote:
On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
First get it working and backup
To get it working you need to run fsck from the rescue media on the link above.
It is a powerful rescue image, not install media.
Does it work on a 15.2 system?
Certainly. You are only doing an fsck. But if that worries you, you can use the 15.2 image instead:
<http://download.opensuse.org/distribution/leap/15.2/live/openSUSE-Leap-15.2-... <http://download.opensuse.org/distribution/leap/15.2/live/openSUSE-Leap-15.2-Rescue-CD-x86_64-Build31.632-Media.iso>>
Are their instructions on the disk because the 15.2 rescue / installdisk had none?
Instructions?
It is just a Linux live, complete, graphical mode, xfce. You are going to do "fsck /dev/sdXY" in a terminal. You have "man fsck". You have internet, mail, browser...
The problem is graphical mode does not come up in Knoppix.
Why are you using knopix instead of the openSUSE rescue image I told you to use?
I am beginning to suspect its the graphics chip. The final test will be a 15.5 install attempt after I finish backing up to external EXT4 based drive.
I can not get the konsole mode of knoppix to recognize the hd on the laptop.
CWSIV
-- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
From: Carl Spitzer {L Juno} <lynux@juno.com> Date: Sun, 23 Jul 2023 02:59:54 -0700 On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote: . . . Thought it was the surge protector going out. But that would not affect booting up and the other system is on it and ok. Ah! -- if you have another system you can boot into, you should be able to do the backup and fsck from there, right? Not thinking straight lost two cats to disease one cancer the other FIV. Also loosing my 15.2 install disk is also not a good idea. Having the exact version is helpful, but not always necessary, especially for ext4, which is quite stable. Other versions would be fine for doing backups. Complete shut down. Only happens in FF and twice in one day. Very weird; FF should not be able to do that. I think its a FF memory management issue because both times I had >15 tabs open. Even so, the OOM killer should be able to keep Linux up and running. ================ From: "Carlos E. R." <robin.listas@telefonica.net> Date: Sun, 23 Jul 2023 12:20:13 +0200 On 2023-07-23 11:59, Carl Spitzer {L Juno} wrote:
On Sat, 2023-07-22 at 22:32 -0500, David C. Rankin wrote:
On 7/22/23 22:14, Carl Spitzer {L Juno} wrote:
Thankfully I have another system old 32 bit but I can download and burn dvd from it.
Then download the rescue CD, and put it in an USB, not a CD. Then the rescue live can write files, even install things that you miss . . . If for some reason you can't use the other system on the disk, then do what Carlos says. -- Bob
Hello, In the Message; Subject : Re: Corrupted root Message-ID : <1690106394.19908.6.camel@linux-7k5b.site> Date & Time: Sun, 23 Jul 2023 02:59:54 -0700 [CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written: [...] CS> Complete shut down. Only happens in FF and twice in one day. CS> I think its a FF memory management issue because both times I had >15 CS> tabs open. Opening a tab consumes a corresponding amount of memory, which you can easily ascertain this with the free command. Your upgrade to Leap 15.5 will not change this situation. So, if you want to use FF with many tabs open (which is certainly convenient), you will need adding more memory. Regards. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "No Windows, no gains!" ... "Why, I am wrong?" -- Bill --
On Mon, 2023-07-24 at 09:28 +0900, Masaru Nomiya wrote:
Hello,
In the Message;
Subject : Re: Corrupted root Message-ID : <1690106394.19908.6.camel@linux-7k5b.site> Date & Time: Sun, 23 Jul 2023 02:59:54 -0700
[CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written:
[...] CS> Complete shut down. Only happens in FF and twice in one day. CS> I think its a FF memory management issue because both times I had >15 CS> tabs open.
Opening a tab consumes a corresponding amount of memory, which you can easily ascertain this with the free command.
Your upgrade to Leap 15.5 will not change this situation.
So, if you want to use FF with many tabs open (which is certainly convenient), you will need adding more memory.
The Dell n5110 is maxed out on memory. I will just have to manage FF better. Shame is they do not fix the problem so when it crashes it does not take down a system. It happened last time I used Ubuntu and FF. So it definitely is a problem. Thankfully there is still Seamonkey and Opera for other things. CWSIV
Hello, In the Message; Subject : Re: Corrupted root Message-ID : <1690161123.28348.4.camel@linux-7k5b.site> Date & Time: Sun, 23 Jul 2023 18:12:03 -0700 [CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written: CS> On Mon, 2023-07-24 at 09:28 +0900, Masaru Nomiya wrote: [...] MN> > Opening a tab consumes a corresponding amount of memory, which you can MN> > easily ascertain this with the free command. MN> > MN> > Your upgrade to Leap 15.5 will not change this situation. MN> > MN> > So, if you want to use FF with many tabs open (which is certainly MN> > convenient), you will need adding more memory. CS> The Dell n5110 is maxed out on memory. Is it so! CS> I will just have to manage FF better. CS> Shame is they do not fix the problem so when it crashes it does CS> not take down a system. CS> It happened last time I used Ubuntu and FF. So it definitely is CS> a problem. Many reports of FF crashes. Rather, there are too many! The developers are aware of this and have shown countermeasures, but ...... CS> Thankfully there is still Seamonkey and Opera for other things. You can use David's recommended Opera.... Regards. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "Maddox hopes that empowering users to pick their own algorithms will get them to think more about what’s involved in making them. " -- Bluesky's Custom Algorithms Could Be the Future of Social Media --
On 2023-07-24 03:12, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 09:28 +0900, Masaru Nomiya wrote:
Hello,
In the Message;
Subject : Re: Corrupted root Message-ID : <1690106394.19908.6.camel@linux-7k5b.site <mailto:1690106394.19908.6.camel@linux-7k5b.site>> Date & Time: Sun, 23 Jul 2023 02:59:54 -0700
[CS] == Carl Spitzer {L Juno} <lynux@juno.com <mailto:lynux@juno.com>> has written:
[...] CS> Complete shut down. Only happens in FF and twice in one day. CS> I think its a FF memory management issue because both times I had >15 CS> tabs open.
Opening a tab consumes a corresponding amount of memory, which you can easily ascertain this with the free command.
Your upgrade to Leap 15.5 will not change this situation.
So, if you want to use FF with many tabs open (which is certainly convenient), you will need adding more memory.
The Dell n5110 is maxed out on memory. I will just have to manage FF better. Shame is they do not fix the problem so when it crashes it does not take down a system. It happened last time I used Ubuntu and FF. So it definitely is a problem.
That is peculiar to you, other people don't have that problem. My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down. Do you have enough Swap space? I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
Le 24/07/2023 à 12:46, Carlos E. R. a écrit :
On 2023-07-24 03:12, Carl Spitzer {L Juno} wrote:
The Dell n5110 is maxed out on memory.
I didn't read all, what is the actal memory? run "free" on a terminal and paste the result here, this will also give the swap size jdd -- c'est quoi, usenet? http://www.dodin.org/wiki/pmwiki.php?n=Usenet.Usenet
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap. What would a graphics card failure look like?? Knoppix in two versions failed to come up 8.6.1 and 9.1?? CWSIV
On Jul 24, 2023, at 9:04 AM, Carl Spitzer {L Juno} <lynux@juno.com> wrote:
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap.
What would a graphics card failure look like?? Knoppix in two versions failed to come up 8.6.1 and 9.1??
CWSIV
This has all the signs of a memory stick failing. If you multiple sticks remove all of the and run on one at a time to try and isolate which one may be bad. I had this same problem in my tower and quickly found the faulty stick. I replaced it and have not had a problem since. Ken
Le 24/07/2023 à 16:13, kschneider bout-tyme.net a écrit :
This has all the signs of a memory stick failing.
if you can boot, run "memtest86", available on most repair disk and you'll know jd -- c'est quoi, usenet? http://www.dodin.org/wiki/pmwiki.php?n=Usenet.Usenet
On Mon, 2023-07-24 at 14:13 +0000, kschneider bout-tyme.net wrote:
On Jul 24, 2023, at 9:04 AM, Carl Spitzer {L Juno} <lynux@juno.com> wrote:
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap.
What would a graphics card failure look like?? Knoppix in two versions failed to come up 8.6.1 and 9.1??
CWSIV
This has all the signs of a memory stick failing. If you multiple sticks remove all of the and run on one at a time to try and isolate which one may be bad. I had this same problem in my tower and quickly found the faulty stick. I replaced it and have not had a problem since.
Ken
On a laptop that is major surgery. Right now its up in konsole mode so I am backing up the /home. CWSIV
On 2023-07-24 15:01, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap.
Small. Well, you tell us. Run "free" or "free -h" as we asked, when firefox is loaded. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On 7/24/23 10:37, Carlos E. R. wrote:
As I remember 2GB for swap.
Small.
Depends on the machine/load. Granted I tend to run boinc at 100% 24x7, but opihi is my desktop machine (both 15.5). bill@opihi:~> uptime ; free -h 10:57:55 up 2:04, 10 users, load average: 16.66, 16.45, 16.42 total used free shared buff/cache available Mem: 30Gi 7.7Gi 20Gi 79Mi 2.6Gi 22Gi Swap: 2.0Gi 0B 2.0Gi bill@kraken:~> uptime ; free -h 10:58:15 up 2:08, 1 user, load average: 33.48, 33.45, 33.39 total used free shared buff/cache available Mem: 31Gi 3.5Gi 27Gi 21Mi 927Mi 27Gi Swap: 2.0Gi 0B 2.0Gi
On 2023-07-24 21:05, Bill Swisher wrote:
On 7/24/23 10:37, Carlos E. R. wrote:
As I remember 2GB for swap.
Small.
Depends on the machine/load. Granted I tend to run boinc at 100% 24x7, but opihi is my desktop machine (both 15.5).
bill@opihi:~> uptime ; free -h 10:57:55 up 2:04, 10 users, load average: 16.66, 16.45, 16.42 total used free shared buff/cache available Mem: 30Gi 7.7Gi 20Gi 79Mi 2.6Gi 22Gi Swap: 2.0Gi 0B 2.0Gi
bill@kraken:~> uptime ; free -h 10:58:15 up 2:08, 1 user, load average: 33.48, 33.45, 33.39 total used free shared buff/cache available Mem: 31Gi 3.5Gi 27Gi 21Mi 927Mi 27Gi Swap: 2.0Gi 0B 2.0Gi
Yes, it depends on the machine/load. My guess is it will be high for Carl, if firefox is running for a while. Notice he has 8GiB ram, you have over 30. cer@Telcontar:~> uptime ; free -h 22:23:47 up 100 days 22:41, 2 users, load average: 1,84, 1,87, 1,87 total used free shared buff/cache available Mem: 62Gi 36Gi 19Gi 186Mi 7,3Gi 26Gi Swap: 99Gi 17Gi 82Gi cer@Telcontar:~> But this machine uses hibernation. My "server": Isengard:~ # uptime ; free -h 22:24:41 up 62 days 16:30, 6 users, load average: 2.67, 2.80, 2.65 total used free shared buff/cache available Mem: 7.6Gi 3.4Gi 155Mi 129Mi 4.5Gi 4.2Gi Swap: 9.0Gi 1.7Gi 7.3Gi Isengard:~ # The big memory user here is firefox, which I use to display youtubes. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On Mon, 2023-07-24 at 20:37 +0200, Carlos E. R. wrote:
On 2023-07-24 15:01, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap.
Small.
Well, you tell us. Run "free" or "free -h" as we asked, when firefox is loaded.
As soon as graphical mode comes up. First finish backup then attempt 15.5 install. Since Knopix does not come up into graphical mode it may be time for a new laptop. CWSIV
On 2023-07-25 03:02, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 20:37 +0200, Carlos E. R. wrote:
On 2023-07-24 15:01, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap.
Small.
Well, you tell us. Run "free" or "free -h" as we asked, when firefox is loaded.
As soon as graphical mode comes up. First finish backup then attempt 15.5 install.
Since Knopix does not come up into graphical mode it may be time for a new laptop.
Why are you not using the opensuse image I told you to use instead of Knopix? -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On Tue, 2023-07-25 at 12:37 +0200, Carlos E. R. wrote:
On 2023-07-25 03:02, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 20:37 +0200, Carlos E. R. wrote:
On 2023-07-24 15:01, Carl Spitzer {L Juno} wrote:
On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote:
That is peculiar to you, other people don't have that problem.
My guess is that you have an OOM situation, and the kernel starts killing things and bringing the system down.
Do you have enough Swap space?
I have, for instance, something like 80 firefox windows and uncountable tabs. Sometimes, 3 firefox instances. No crashes, uptime is 100 days.
As I remember 2GB for swap.
Small.
Well, you tell us. Run "free" or "free -h" as we asked, when firefox is loaded.
As soon as graphical mode comes up. First finish backup then attempt 15.5 install.
Since Knopix does not come up into graphical mode it may be time for a new laptop.
Why are you not using the opensuse image I told you to use instead of Knopix?
It did not work. Remember its the root. I kept getting a fail on umount for the root because it was busy. I tried with the rescue somehow either I forgot how or it did not work. Knoppix has always been a backup test for any system, until now. I think the Knoppix 9.1 vintage 2021 did not accept the NVIDIA graphics yet I know the system is older than that. CWSIV
Hello, In the Message; Subject : Re: Corrupted root Message-ID : <1690203699.3060.6.camel@linux-7k5b.site> Date & Time: Mon, 24 Jul 2023 06:01:39 -0700 [CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written: CS> [1 <text/plain; UTF-8 (7bit)>] CS> On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote: [...] MN> > That is peculiar to you, other people don't have that problem. CS> As I remember 2GB for swap. Now I know why FF crashes. I thought it was because it's a notebook PC, so it doesn't have that much memory (8GB?). I thought that was the cause of the FF crash, but it is not that, but too small swap size. The swap size should be set to the memory + 2GB (probably 10GB?). Even if you set it that way, beware of FF use. Regards. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "Maddox hopes that empowering users to pick their own algorithms will get them to think more about what’s involved in making them. " -- Bluesky's Custom Algorithms Could Be the Future of Social Media --
On Tue, 25 Jul 2023 09:11:14 +0900 Masaru Nomiya <nomiya@lake.dti.ne.jp> wrote:
Hello,
In the Message;
Subject : Re: Corrupted root Message-ID : <1690203699.3060.6.camel@linux-7k5b.site> Date & Time: Mon, 24 Jul 2023 06:01:39 -0700
[CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written:
CS> [1 <text/plain; UTF-8 (7bit)>] CS> On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote: [...] MN> > That is peculiar to you, other people don't have that problem.
CS> As I remember 2GB for swap.
Now I know why FF crashes.
I thought it was because it's a notebook PC, so it doesn't have that much memory (8GB?). I thought that was the cause of the FF crash, but it is not that, but too small swap size.
The swap size should be set to the memory + 2GB (probably 10GB?).
Do you have a reference to support that assertion, please? (preferably an authoritative reference :)
Even if you set it that way, beware of FF use.
On 2023-07-25 11:24, Dave Howorth wrote:
On Tue, 25 Jul 2023 09:11:14 +0900 Masaru Nomiya <> wrote:
Hello,
In the Message;
Subject : Re: Corrupted root Message-ID : <1690203699.3060.6.camel@linux-7k5b.site> Date & Time: Mon, 24 Jul 2023 06:01:39 -0700
[CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written:
CS> [1 <text/plain; UTF-8 (7bit)>] CS> On Mon, 2023-07-24 at 12:46 +0200, Carlos E. R. wrote: [...] MN> > That is peculiar to you, other people don't have that problem.
CS> As I remember 2GB for swap.
Now I know why FF crashes.
I thought it was because it's a notebook PC, so it doesn't have that much memory (8GB?). I thought that was the cause of the FF crash, but it is not that, but too small swap size.
The swap size should be set to the memory + 2GB (probably 10GB?).
Do you have a reference to support that assertion, please? (preferably an authoritative reference :)
For a laptop, yes, recommended swap, probably automatically by YaST, is memory size and a bit more ;-D My own recommendation for a computer with 8 GiB of RAM with heavy use of Firefox would be 12..16 GiB of SWAP in an SSD ;-) -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
Hello, In the Message; Subject : Re: Corrupted root Message-ID : <e3913bff-1048-6e9d-7b00-deb8252c158b@telefonica.net> Date & Time: Tue, 25 Jul 2023 12:41:34 +0200 [CER] == "Carlos E. R." <robin.listas@telefonica.net> has written: CER> On 2023-07-25 11:24, Dave Howorth wrote: CER> > On Tue, 25 Jul 2023 09:11:14 +0900 DH> > Masaru Nomiya <> wrote: [...] MN>>> Now I know why FF crashes. MN>>> I thought it was because it's a notebook PC, so it doesn't have that MN>>> much memory (8GB?). I thought that was the cause of the FF crash, but MN>>> it is not that, but too small swap size. MN>>> The swap size should be set to the memory + 2GB (probably 10GB?). DH>> Do you have a reference to support that assertion, please? DH>> (preferably an authoritative reference :) CER> For a laptop, yes, recommended swap, probably automatically by CER> YaST, is memory size and a bit more ;-D CER> My own recommendation for a computer with 8 GiB of RAM with CER> heavy use of Firefox would be 12..16 GiB of SWAP in an SSD ;-) Thanks, Carlos. The opinion that the conventional argument about swap does not apply to SSDs because their transfer rate is much faster than that of HDDs is presented. But this is questionable, as Carlos also implies. That is, I am using Tumbleweed on a 4TB SDD and 96GB memory PC with 4GB swap, and when I open as many Tabs as Carl does in vivaldi, swap is consumed as I see it. So I sometimes run # swapoff -a && swapon -a. The transfer rate has anything to do with the swap size, I wonder? Anyway, as for David's question, the reference is a well-known Japanese site for Japanese Linux users, which is written based on kernel developer Chris Down's answer to the Japanese user's question, "Is swap necessary in the current PC's situation with a lot of memory?". https://chrisdown.name/2018/01/02/in-defence-of-swap.html Based on this, it is said that the standard swap size is the amount of installed memory(in case of < 32GB) + 2GB since kernel 4.0, and I think that YaST2 also decides the swap size according to the same idea. Needless to say, this does not take into account hibernation, which Carlos often mentions, with considering that Carl does not use it. Regards & Good Night. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "Maddox hopes that empowering users to pick their own algorithms will get them to think more about what’s involved in making them. " -- Bluesky's Custom Algorithms Could Be the Future of Social Media --
On 2023-07-25 15:02, Masaru Nomiya wrote:
Hello,
In the Message;
Subject : Re: Corrupted root Message-ID : <e3913bff-1048-6e9d-7b00-deb8252c158b@telefonica.net> Date & Time: Tue, 25 Jul 2023 12:41:34 +0200
[CER] == "Carlos E. R." <robin.listas@telefonica.net> has written:
CER> On 2023-07-25 11:24, Dave Howorth wrote: CER> > On Tue, 25 Jul 2023 09:11:14 +0900 DH> > Masaru Nomiya <> wrote: [...] MN>>> Now I know why FF crashes.
MN>>> I thought it was because it's a notebook PC, so it doesn't have that MN>>> much memory (8GB?). I thought that was the cause of the FF crash, but MN>>> it is not that, but too small swap size.
MN>>> The swap size should be set to the memory + 2GB (probably 10GB?).
DH>> Do you have a reference to support that assertion, please? DH>> (preferably an authoritative reference :)
CER> For a laptop, yes, recommended swap, probably automatically by CER> YaST, is memory size and a bit more ;-D
CER> My own recommendation for a computer with 8 GiB of RAM with CER> heavy use of Firefox would be 12..16 GiB of SWAP in an SSD ;-)
Thanks, Carlos.
The opinion that the conventional argument about swap does not apply to SSDs because their transfer rate is much faster than that of HDDs is presented.
But this is questionable, as Carlos also implies.
I do not recommend using a "rotating rust" hard disk if swap is in actual use. For hibernation it is ok. The reason is that I observed a huge slowdown with swap on HD I don't remember on what version of openSUSE. From one version to the next speed dropped a lot. I blamed it on "fragmentation" of the memory space on swap. There was a lot of head movement, and performance is, or was terrible. The computer was barely usable. Then I put swap on SSD, and the computer became usable again (an 8 GiB computer with heavy usage of Firefox), years ago. I still use that machine on an alternate location.
That is, I am using Tumbleweed on a 4TB SDD and 96GB memory PC with 4GB swap, and when I open as many Tabs as Carl does in vivaldi, swap is consumed as I see it. So I sometimes run # swapoff -a && swapon -a.
You do not gain speed doing that, IMO.
The transfer rate has anything to do with the swap size, I wonder?
I don't think swap size has effect on transfer rate.
Anyway, as for David's question, the reference is a well-known Japanese site for Japanese Linux users, which is written based on kernel developer Chris Down's answer to the Japanese user's question, "Is swap necessary in the current PC's situation with a lot of memory?".
https://chrisdown.name/2018/01/02/in-defence-of-swap.html
Based on this, it is said that the standard swap size is the amount of installed memory(in case of < 32GB) + 2GB since kernel 4.0, and I think that YaST2 also decides the swap size according to the same idea.
Needless to say, this does not take into account hibernation, which Carlos often mentions, with considering that Carl does not use it.
Yast probably considers that a laptop should have enough swap for hibernation, as a default choice. I can't confirm, though. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
Hello, Sorry, I missed. In the Message; Subject : Re: Corrupted root Message-ID : <4c3cca0b-d67f-51da-0e0b-9f6c4d9dde1d@telefonica.net> Date & Time: Tue, 25 Jul 2023 15:14:46 +0200 [CER] == "Carlos E. R." <robin.listas@telefonica.net> has written: [...] MN> > The opinion that the conventional argument about swap does not apply MN> > to SSDs because their transfer rate is much faster than that of HDDs MN> > is presented. MN> > MN> > But this is questionable, as Carlos also implies. CER> I do not recommend using a "rotating rust" hard disk if swap is in actual CER> use. For hibernation it is ok. [...] It was a long time ago, so I don't remember exactly, but what I saw was an opinion that when using SSD, we can set a smaller swap than HDD. MN> > That is, I am using Tumbleweed on a 4TB SDD and 96GB memory PC with MN> > 4GB swap, and when I open as many Tabs as Carl does in vivaldi, swap MN> > is consumed as I see it. So I sometimes run # swapoff -a && swapon -a. CER> You do not gain speed doing that, IMO. The reason is not speed, but to free up more swap space. Now that I know how to use brave with Enlightenment, I stopped using the swap-eating vivaldi and I'm good to go. Thanks David for this point! Regards. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "Maddox hopes that empowering users to pick their own algorithms will get them to think more about what’s involved in making them. " -- Bluesky's Custom Algorithms Could Be the Future of Social Media --
On 2023-07-28 11:24, 野宮 賢 / NOMIYA Masaru wrote:
Hello,
Sorry, I missed.
In the Message;
Subject : Re: Corrupted root Message-ID : <4c3cca0b-d67f-51da-0e0b-9f6c4d9dde1d@telefonica.net> Date & Time: Tue, 25 Jul 2023 15:14:46 +0200
[CER] == "Carlos E. R." <robin.listas@telefonica.net> has written:
[...] MN> > The opinion that the conventional argument about swap does not apply MN> > to SSDs because their transfer rate is much faster than that of HDDs MN> > is presented. MN> > MN> > But this is questionable, as Carlos also implies.
CER> I do not recommend using a "rotating rust" hard disk if swap is in actual CER> use. For hibernation it is ok. [...]
It was a long time ago, so I don't remember exactly, but what I saw was an opinion that when using SSD, we can set a smaller swap than HDD.
:-? Strange.
MN> > That is, I am using Tumbleweed on a 4TB SDD and 96GB memory PC with MN> > 4GB swap, and when I open as many Tabs as Carl does in vivaldi, swap MN> > is consumed as I see it. So I sometimes run # swapoff -a && swapon -a.
CER> You do not gain speed doing that, IMO.
The reason is not speed, but to free up more swap space.
In that situation, I would instead restart the app.
Now that I know how to use brave with Enlightenment, I stopped using the swap-eating vivaldi and I'm good to go.
Thanks David for this point!
-- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
Hello, In the Message; Subject : Re: Corrupted root Message-ID : <7da469a4-d07c-0af0-c55a-7b8d1eb541db@telefonica.net> Date & Time: Fri, 28 Jul 2023 12:01:06 +0200 [CER] == "Carlos E. R." <robin.listas@telefonica.net> has written: CER> On 2023-07-28 11:24, 野宮 賢 / NOMIYA Masaru wrote: [...] MN>> It was a long time ago, so I don't remember exactly, but what I saw MN>> was an opinion that when using SSD, we can set a smaller swap than MN>> HDD. CER> :-? CER> Strange. I agree. [...] CER>>> You do not gain speed doing that, IMO. MN>> The reason is not speed, but to free up more swap space. CER> In that situation, I would instead restart the app. No, the real solution is logout && login. In other words, swapoff -a && swapon -a often does not lighten the load. Brave is very comfortable, by the way. Reards. --- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "Maddox hopes that empowering users to pick their own algorithms will get them to think more about what’s involved in making them. " -- Bluesky's Custom Algorithms Could Be the Future of Social Media --
On 2023-07-28 12:25, Masaru Nomiya wrote:
Hello,
In the Message;
Subject : Re: Corrupted root Message-ID : <7da469a4-d07c-0af0-c55a-7b8d1eb541db@telefonica.net> Date & Time: Fri, 28 Jul 2023 12:01:06 +0200
[CER] == "Carlos E. R." <robin.listas@telefonica.net> has written:
CER> On 2023-07-28 11:24, 野宮 賢 / NOMIYA Masaru wrote:
[...] MN>> It was a long time ago, so I don't remember exactly, but what I saw MN>> was an opinion that when using SSD, we can set a smaller swap than MN>> HDD.
CER> :-?
CER> Strange.
I agree.
[...] CER>>> You do not gain speed doing that, IMO.
MN>> The reason is not speed, but to free up more swap space.
CER> In that situation, I would instead restart the app.
No, the real solution is logout && login.
When a browser eats memory and ends swapping, just restarting it works nicely :-)
In other words, swapoff -a && swapon -a often does not lighten the load.
Brave is very comfortable, by the way.
-- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On Tue, 25 Jul 2023 22:02:55 +0900 Masaru Nomiya <nomiya@lake.dti.ne.jp> wrote:
Hello, [snip] MN>>> The swap size should be set to the memory + 2GB (probably MN>>> 10GB?).
DH>> Do you have a reference to support that assertion, please? DH>> (preferably an authoritative reference :) [snip] Anyway, as for David's question, the reference is a well-known Japanese site for Japanese Linux users, which is written based on kernel developer Chris Down's answer to the Japanese user's question, "Is swap necessary in the current PC's situation with a lot of memory?".
https://chrisdown.name/2018/01/02/in-defence-of-swap.html
Based on this, it is said that the standard swap size is the amount of installed memory(in case of < 32GB) + 2GB since kernel 4.0, and I think that YaST2 also decides the swap size according to the same idea.
Many thanks :)
Needless to say, this does not take into account hibernation, which Carlos often mentions, with considering that Carl does not use it.
Regards & Good Night.
--- ┏━━┓彡 野宮 賢 mail-to: nomiya @ lake.dti.ne.jp ┃＼／彡 ┗━━┛ "Maddox hopes that empowering users to pick their own algorithms will get them to think more about what’s involved in making them. " -- Bluesky's Custom Algorithms Could Be the Future of Social Media --
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine) On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM. Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address.... -- David C. Rankin, J.D.,P.E.
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap. Jim F
On Tue, 2023-08-01 at 21:07 -0500, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
I do not think its new enough to be an ssd. for all the writes I do downloading things I would never want an SSD except for backup since they are glorified flash drives. CWSIV
On 8/2/23 00:13, Carl Spitzer {L Juno} wrote:
On Tue, 2023-08-01 at 21:07 -0500, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
I do not think its new enough to be an ssd. for all the writes I do downloading things I would never want an SSD except for backup since they are glorified flash drives.
Say what you will about SSD, but from personal experience they're okay. I've been using them for system disks, including swap, in servers and desktops since approximately 2006 without issue. At any given time I've got about 30 hosts up and running, most with SSD installed. If I remember correctly, I did have one fail, but that's still a good record. I've seen spinners fail at a much greater rate. Indeed, the first desktop I tried them in in 2006 is still up and running. I also use swap configured just a bit bigger than the installed RAM, except on servers with 512-GB of RAM. Back in the SunOS days the rule-of-thumb was to configure swap to be three-times larger than your RAM, but that was then! YMMV of course, but for me SSDs are okay. Regards, Lew
On Wed, 2023-08-02 at 06:50 -0700, Lew Wolfgang wrote:
On 8/2/23 00:13, Carl Spitzer {L Juno} wrote:
On Tue, 2023-08-01 at 21:07 -0500, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
I do not think its new enough to be an ssd. for all the writes I do downloading things I would never want an SSD except for backup since they are glorified flash drives.
Say what you will about SSD, but from personal experience they're okay. I've been using them for system disks, including swap, in servers and desktops since approximately 2006 without issue. At any given time I've got about 30 hosts up and running, most with SSD installed. If I remember correctly, I did have one fail, but that's still a good record. I've seen spinners fail at a much greater rate. Indeed, the first desktop I tried them in in 2006 is still up and running.
I also use swap configured just a bit bigger than the installed RAM, except on servers with 512-GB of RAM. Back in the SunOS days the rule-of-thumb was to configure swap to be three-times larger than your RAM, but that was then!
YMMV of course, but for me SSDs are okay.
What Brand?? Mostly they are much more expensive than normal drives and I do not have a color printer.. CWSIV
On 2023-08-02 17:15, Carl Spitzer {L Juno} wrote:
On Wed, 2023-08-02 at 06:50 -0700, Lew Wolfgang wrote:
On 8/2/23 00:13, Carl Spitzer {L Juno} wrote:
On Tue, 2023-08-01 at 21:07 -0500, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
I do not think its new enough to be an ssd. for all the writes I do downloading things I would never want an SSD except for backup since they are glorified flash drives.
Say what you will about SSD, but from personal experience they're okay. I've been using them for system disks, including swap, in servers and desktops since approximately 2006 without issue. At any given time I've got about 30 hosts up and running, most with SSD installed. If I remember correctly, I did have one fail, but that's still a good record. I've seen spinners fail at a much greater rate. Indeed, the first desktop I tried them in in 2006 is still up and running.
I also use swap configured just a bit bigger than the installed RAM, except on servers with 512-GB of RAM. Back in the SunOS days the rule-of-thumb was to configure swap to be three-times larger than your RAM, but that was then!
YMMV of course, but for me SSDs are okay.
What Brand??
Mostly they are much more expensive than normal drives and I do not have a color printer..
If you buy them big, yes, they are expensive. But for the system half a terabyte is enough. You can have another big rotating rust disk for the rest or just /home. Even 250 GB is enough, I have a machine like that, then 2TB on rotating rust (not /home, but /extra). It is different if you use a laptop, of course. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
On 8/2/23 08:15, Carl Spitzer {L Juno} wrote:
On Wed, 2023-08-02 at 06:50 -0700, Lew Wolfgang wrote:
On 8/2/23 00:13, Carl Spitzer {L Juno} wrote:
On Tue, 2023-08-01 at 21:07 -0500, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
I do not think its new enough to be an ssd. for all the writes I do downloading things I would never want an SSD except for backup since they are glorified flash drives.
Say what you will about SSD, but from personal experience they're okay. I've been using them for system disks, including swap, in servers and desktops since approximately 2006 without issue. At any given time I've got about 30 hosts up and running, most with SSD installed. If I remember correctly, I did have one fail, but that's still a good record. I've seen spinners fail at a much greater rate. Indeed, the first desktop I tried them in in 2006 is still up and running.
I also use swap configured just a bit bigger than the installed RAM, except on servers with 512-GB of RAM. Back in the SunOS days the rule-of-thumb was to configure swap to be three-times larger than your RAM, but that was then!
YMMV of course, but for me SSDs are okay.
What Brand??
Mostly they are much more expensive than normal drives and I do not have a color printer..
For the most part, whatever is available. Probably mostly Intel, second would be Samsung. Yes, they're more expensive, which is why I use them mostly for the root partition. Spinners are mostly used for /home. As such, 256-GB is more than big enough for most root partitions. My home desktop has one 1-TB nvme drive, and two Seagate spinners, 10 and 16 TB. At work I've got one server with two 1-T nvme's configured as a RAID-0 stripe, with 658-TB of RAID-6 Seagate spinners. smartctl reports about 30,000 power-on hours with no errors and 100% of spare sectors available on the nvme's. Regards, Lew
On 2023-08-02 04:07, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
Hardly. On a modern SSD disk, the firmware would simply remap to other sectors, and the disk would behave normally till exhausted. It would show in the SMART tests, which everybody does, right? >:-) I have two machines with swap on SSD, and the disks hardly show any wear after several years of use. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
* Carlos E. R. <robin.listas@telefonica.net> [08-02-23 04:58]:
On 2023-08-02 04:07, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Complete shut down. Only happens in FF and twice in one day. I think its a FF memory management issue because both times I had >15 tabs open.
There is something going on here. I'll run up to 15 or so tabs open and I've never had an issue (some JS demands excess CPU, but memory is fine)
On the install disk, before you go to sleep, start memtest86+ and let it run. It also sounds like you are running into a bad-memory issue. When that occurs -- "is like a box of chocolates", but when you it the bad address, the system will crash - and it may bring the entire system down. With memory address radomization in the Linux virtual memory manager -- there is no telling when, or if, any one program will have access to the bad RAM.
Firefox, in requesting (and in using in some cases) gigabytes of memory -- is one of the more likely apps to hit the bad address....
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
Hardly.
On a modern SSD disk, the firmware would simply remap to other sectors, and the disk would behave normally till exhausted. It would show in the SMART tests, which everybody does, right? >:-)
I have two machines with swap on SSD, and the disks hardly show any wear after several years of use.
yes, I have one >12 years which contains tmp, var, and .cache and have not encountered any noticable errors. -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri Photos: http://wahoo.no-ip.org/piwigo paka @ IRCnet oftc
On 2023-08-02 15:16, Patrick Shanahan wrote:
* Carlos E. R. <> [08-02-23 04:58]:
On 2023-08-02 04:07, Jim Flanagan wrote:
On 7/23/23 10:38 PM, David C. Rankin wrote:
On 7/23/23 04:59, Carl Spitzer {L Juno} wrote:
Another issue could be swap on an SSD. We use them because they are fast. But SSDs have limited write cycles, they get burned out after so many writes. You could be hitting some burned out sectors on the Swap.
Hardly.
On a modern SSD disk, the firmware would simply remap to other sectors, and the disk would behave normally till exhausted. It would show in the SMART tests, which everybody does, right? >:-)
I have two machines with swap on SSD, and the disks hardly show any wear after several years of use.
yes, I have one >12 years which contains tmp, var, and .cache and have not encountered any noticable errors.
My oldest SSD is in my old laptop. 500 gigs. 5739 hours of use, says the disk. Smartctl has this parameter:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
202 Percent_Lifetime_Remain 0x0030 099 099 001 Old_age Offline - 1
I don't know if there is a better command (there is nvme, but doesn't apply to SATA SSDs like this one). I hibernate the machine between uses, which now is only to watch videos in front of the static bike. It's battery is dead.
minas-tirith:~ # free -h total used free shared buff/cache available Mem: 3.7Gi 1.4Gi 1.4Gi 104Mi 1.6Gi 2.4Gi Swap: 6.0Gi 949Mi 5.1Gi minas-tirith:~ #
I have another old machine that swaps on SSD and was intensely used, but I can not access it now. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
From: Carl Spitzer {L Juno} <lynux@juno.com> Date: Sat, 22 Jul 2023 20:14:11 -0700 IN OpenSuse 15.2 I need to umount the / partition but I am getting device busy errors which block umount. Its Ext4 so I should be able to fsck this problem once unmounted. Then I can finish backing up the /home and proceed to upgrade to 15.5. How do I correct this device busy problem? The system crashed and shut off during a Firefox secession I think I had too many tabs open. CWSIV I'm with David here; I don't see how you can umount / after booting -- or even remount it read-only. You also don't say much about the present state of your system (rebooted? just the session crashed? something else?), so it's hard to be clear on what your options are. But if you have no more user sessions, then it may not matter: You may be able to get a clean backup of /home from where you are now, and then just reboot into a rescue system (or anything that doesn't require mounting /) so you can fix /. -- Bob Rogers http://www.rgrjr.com/
Hello, David & Bob. In the Message; Subject : Corrupted root Message-ID : <25788.43553.278475.807670@orion.rgrjr.com> Date & Time: Sat, 22 Jul 2023 21:18:41 -0700 [BR] == Bob Rogers <rogers@rgrjr.com> has written: BR> From: Carl Spitzer {L Juno} <lynux@juno.com> BR> Date: Sat, 22 Jul 2023 20:14:11 -0700 [...] BR> I'm with David here; I don't see how you can umount / after booting -- BR> or even remount it read-only. You also don't say much about the present BR> state of your system (rebooted? just the session crashed? something BR> else?), so it's hard to be clear on what your options are. But if you BR> have no more user sessions, then it may not matter: You may be able to BR> get a clean backup of /home from where you are now, and then just reboot BR> into a rescue system (or anything that doesn't require mounting /) so BR> you can fix /. I agree the problem with Carl's question is, as you say, that it leaves out the important question of under what circumstances. But, is it necessary to run fsck on / that Carl is trying to adopt? In the Message; Subject : Corrupted root Message-ID : <1690082052.15872.3.camel@linux-7k5b.site> Date & Time: Sat, 22 Jul 2023 20:14:11 -0700 [CS] == Carl Spitzer {L Juno} <lynux@juno.com> has written: [...] CS> The system crashed and shut off during a Firefox secession I think CS> I had too many tabs open. This is a typical phenomenon caused by lack of memory, but does this phenomenon cause system corruption? I can't figure this out..... Regards.
On 2023-07-23 05:14, Carl Spitzer {L Juno} wrote:
IN OpenSuse 15.2 I need to umount the / partition but I am getting device busy errors which block umount.
Its Ext4 so I should be able to fsck this problem once unmounted. Then I can finish backing up the /home and proceed to upgrade to 15.5.
How do I correct this device busy problem?
You can not. Instead, boot from rescue system on USB stick or CD, and do the fsck from there. If you don't have a rescue system on stick, create one using this image: <http://download.opensuse.org/distribution/leap/15.5/live/openSUSE-Leap-15.5-Rescue-CD-x86_64-Media.iso> If you are unable to create one, use the installation DVD. Boot, choose rescue. As soon as your system boots, create the rescue stick and keep it safe. I pre-systemd times, creating an empty file "/forcefsck" and booting, would force an autocheck. Try that.
The system crashed and shut off during a Firefox secession I think I had too many tabs open.
I don't see why that would corrupt root, but who knows. -- Cheers / Saludos, Carlos E. R. (from 15.4 x86_64 at Telcontar)
participants (13)
-
Bill Swisher -
Bob Rogers -
Carl Spitzer {L Juno} -
Carlos E. R. -
Dave Howorth -
David C. Rankin -
jdd＠dodin.org -
Jim Flanagan -
kschneider bout-tyme.net -
Lew Wolfgang -
Masaru Nomiya -
Patrick Shanahan -
野宮 賢 / NOMIYA Masaru