[opensuse] PC freezes when transferring huge amount of data to USB
Hi, I am noticing a strange problem. I am running openSUSE 12.1 and when I try to copy data on USB sticks the PC freezes and resumes only when I unplug the USB device. I have tried with couple of devices and the result is same. Swapnil -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Wed, Mar 14, 2012 at 01:11:46PM +0100, Swapnil Bhartiya wrote:
I am noticing a strange problem. I am running openSUSE 12.1 and when I try to copy data on USB sticks the PC freezes and resumes only when I unplug the USB device. I have tried with couple of devices and the result is same.
Try to boot with this additional kernel argument: transparent_hugepage=never In the running system you'll see the change from the content of /sys/kernel/mm/transparent_hugepage/enabled This is more or less a workaroud to a known issue. If this helps in your case please do the following: a) Check bugzilla if there is a bug report regarding this. If ther eis one add the details of your system and setup (kernel flavour in use and so on). b) Report the filed or found bug ID back to this list. I've not seen it on any of my workstations or latops. Thanks, Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
What is the status with this issue? Is it solved or did you file a bug as suggested? Please report back to the list. Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
I rebooted the system and the problem went away. It happens frequently. I don't know if it is caused by updates or what. I will report back or file a bug once the problem recurs. If there is a command that I should run which will give some output that can help identify the problem please tell so that I am ready next time. Since its my production machine, at time it makes more sense for me to just reboot than wait, so if the command is handy I will be able to file the report. Best Thanks Swapnul On 03/15/2012 10:18 AM, Lars Müller wrote:
What is the status with this issue?
Is it solved or did you file a bug as suggested?
Please report back to the list.
Lars
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Thu, Mar 15, 2012 at 11:41:08AM +0100, Swapnil Bhartiya wrote:
I rebooted the system and the problem went away. It happens frequently. I don't know if it is caused by updates or what. I will report back or file a bug once the problem recurs. If there is a command that I should run which will give some output that can help identify the problem please tell so that I am ready next time. Since its my production machine, at time it makes more sense for me to just reboot than wait, so if the command is handy I will be able to file the report.
Have you read my first reply? http://lists.openSUSE.org/opensuse/2012-03/msg00655.html It included a suggested what to test first and how to proceed further. Even if the suggestion helps you the issue has to be reported via bugzilla. Not all developers and package maintainers are reading this list. It's up to you to drive this issue from now on. We all rely on you! Thanks, Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
On 03/15/2012 10:53 AM, Lars Müller wrote:
On Thu, Mar 15, 2012 at 11:41:08AM +0100, Swapnil Bhartiya wrote:
I rebooted the system and the problem went away. It happens frequently. I don't know if it is caused by updates or what. I will report back or file a bug once the problem recurs. If there is a command that I should run which will give some output that can help identify the problem please tell so that I am ready next time. Since its my production machine, at time it makes more sense for me to just reboot than wait, so if the command is handy I will be able to file the report. Have you read my first reply?
http://lists.openSUSE.org/opensuse/2012-03/msg00655.html
It included a suggested what to test first and how to proceed further.
Even if the suggestion helps you the issue has to be reported via bugzilla.
Not all developers and package maintainers are reading this list.
It's up to you to drive this issue from now on. We all rely on you!
Thanks. I will make that a standard practice from now on. :-) Best Swapnil
Thanks,
Lars
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Thu, Mar 15, 2012 at 11:04:27AM +0100, Swapnil Bhartiya wrote:
On 03/15/2012 10:53 AM, Lars Müller wrote:
On Thu, Mar 15, 2012 at 11:41:08AM +0100, Swapnil Bhartiya wrote:
I rebooted the system and the problem went away. It happens frequently. I don't know if it is caused by updates or what. I will report back or file a bug once the problem recurs. If there is a command that I should run which will give some output that can help identify the problem please tell so that I am ready next time.
It's more likely to get such instructions if you file the requested bug report.
Since its my production machine, at time it makes more sense for me to just reboot than wait, so if the command is handy I will be able to file the report.
If it would be that simple and I would know more details about the kernel internals I would have pointed you to it. As that's not the case you must got the buzgzilla route or forget about it. You know what TINA stands for? It's not only a lovely name, it is the abbreviation for There Is No Alternative.
Have you read my first reply?
http://lists.openSUSE.org/opensuse/2012-03/msg00655.html
It included a suggested what to test first and how to proceed further.
Even if the suggestion helps you the issue has to be reported via bugzilla.
Not all developers and package maintainers are reading this list.
It's up to you to drive this issue from now on. We all rely on you!
Thanks. I will make that a standard practice from now on. :-)
A asci art smily doesn't compensate the missing bug report. ;) Would you be this nice to file one or quote the bug URL here? You got straight instructions about what's needed and what to do. See http://lists.openSUSE.org/opensuse/2012-03/msg00657.html Is there anything missing to file a bug report? Or is this mainly a question of time? What ever it is please reply with the information requested with the first reply in this thread. Do this via bugzilla, please. As written before on this list not all people working on the development of openSUSE or SUSE Linux Enterprise follow this list. Thefore please, please use bugzilla. Then you might get the attention of the person in charge for the particular software piece. On Thu, Mar 15, 2012 at 01:18:38PM +0100, Graham Anderson wrote:
If it's not the issue you mentioned, is it possible he is seeing blocking on I/O that many people experience with badly aligned partitions on USB sticks?
Might even be this. But if this happens to other USB mass storage devices too I stronly suggest to go the bugzilla route as suggested before and above. Thanks, Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
On Wednesday 14 Mar 2012 13:37:11 Lars Müller wrote:
On Wed, Mar 14, 2012 at 01:11:46PM +0100, Swapnil Bhartiya wrote:
I am noticing a strange problem. I am running openSUSE 12.1 and when I try to copy data on USB sticks the PC freezes and resumes only when I unplug the USB device. I have tried with couple of devices and the result is same.
Try to boot with this additional kernel argument:
transparent_hugepage=never
In the running system you'll see the change from the content of
/sys/kernel/mm/transparent_hugepage/enabled
This is more or less a workaroud to a known issue.
If this helps in your case please do the following:
a) Check bugzilla if there is a bug report regarding this.
If ther eis one add the details of your system and setup (kernel flavour in use and so on).
b) Report the filed or found bug ID back to this list.
I've not seen it on any of my workstations or latops.
Thanks,
Lars
If it's not the issue you mentioned, is it possible he is seeing blocking on I/O that many people experience with badly aligned partitions on USB sticks? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Content-ID: <alpine.LNX.2.00.1203182222210.953@Telcontar.valinor> On Wednesday, 2012-03-14 at 13:37 +0100, Lars Müller wrote:
Try to boot with this additional kernel argument:
transparent_hugepage=never
Some one in the forum had this problem, and that setting solved it. View this thread: http://forums.opensuse.org/showthread.php?t=473300 I told him to write a bugzilla and tell you, but I'm afraid he doesn't want to. :-( - -- Cheers, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) iEYEARECAAYFAk9mUlkACgkQtTMYHG2NR9VhQACglwRqAL1v3akOlQJtrrc/b7Sp c6gAn07P86OFqgfnANcC9lNI3j7cJ//g =0aFo -----END PGP SIGNATURE-----
On Sun, Mar 18, 2012 at 10:23:27PM +0100, Carlos E. R. wrote:
On Wednesday, 2012-03-14 at 13:37 +0100, Lars Müller wrote:
Try to boot with this additional kernel argument:
transparent_hugepage=never
Some one in the forum had this problem, and that setting solved it.
View this thread: http://forums.opensuse.org/showthread.php?t=473300
I told him to write a bugzilla and tell you, but I'm afraid he doesn't want to. :-(
Look, if someone doesn't see the need to communicate a bug issue the predefined way, then this issue might be less pressing. Also not all Open Source Software developers eager to be subscribed to yet another mailing list. Ask yourself why should a Linux kernel or a glibc guy hang around here? Only a bugreport will get visible to the Linux kernel developers. In this particular question they're aware of it. Unfortunately I don't know a bug ID which is used to track this issue. By filing a new report we at least would soon see if it got reported in the past at all. Recently we had a guy on this or the German list not being able better not being willing to offer the requested information. Even after asking back a second time. And even after giving the detailed description what's need. What are you suggesting to do in such a case? To burn more time with such people? I don't even black list them. I already forgot the name and particular issue we discussed as anything else is a waste of time and energy. I believe all we can do is to enhance the documentation and to make the process of offering the required inforamtion as smooth as possible. We always must keep in mind that there is a small amount of people with only one goal: to get their individual issue solved. They don't care about others nor about the progress of Open Source Software in general. Well, that might work one or two times for them. I expect the majority of people active in the OSS community unmasks such people very fast. I suggest to wait some days and then request the needed information as outlined by http://lists.openSUSE.org/opensuse/2012-03/msg00657.html again. Some people are not this fast or are not aware that their input might help to solve a generic issue. Cheers, Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
On 03/18/2012 11:45 PM, Lars Müller wrote:
On Sun, Mar 18, 2012 at 10:23:27PM +0100, Carlos E. R. wrote:
On Wednesday, 2012-03-14 at 13:37 +0100, Lars Müller wrote:
Try to boot with this additional kernel argument:
transparent_hugepage=never
Some one in the forum had this problem, and that setting solved it.
View this thread: http://forums.opensuse.org/showthread.php?t=473300
I told him to write a bugzilla and tell you, but I'm afraid he doesn't want to. :-(
Look, if someone doesn't see the need to communicate a bug issue the predefined way, then this issue might be less pressing. Also not all Open Source Software developers eager to be subscribed to yet another mailing list. Ask yourself why should a Linux kernel or a glibc guy hang around here?
Only a bugreport will get visible to the Linux kernel developers. In this particular question they're aware of it. Unfortunately I don't know a bug ID which is used to track this issue. By filing a new report we at least would soon see if it got reported in the past at all.
Recently we had a guy on this or the German list not being able better not being willing to offer the requested information. Even after asking back a second time. And even after giving the detailed description what's need.
What are you suggesting to do in such a case? To burn more time with such people? I don't even black list them. I already forgot the name and particular issue we discussed as anything else is a waste of time and energy.
I believe all we can do is to enhance the documentation and to make the process of offering the required inforamtion as smooth as possible.
We always must keep in mind that there is a small amount of people with only one goal: to get their individual issue solved. They don't care about others nor about the progress of Open Source Software in general.
Well, that might work one or two times for them. I expect the majority of people active in the OSS community unmasks such people very fast.
I suggest to wait some days and then request the needed information as outlined by http://lists.openSUSE.org/opensuse/2012-03/msg00657.html again. Some people are not this fast or are not aware that their input might help to solve a generic issue.
In my case as I said that the problem went away after reboot. I had not rebooted the system for days. I said I would file the report when the problem recurs. Not that I don't want to file it (it's only going to help users like me) but I don't know what to file unless I get the problem again. If just creating a bug report that my system freezes when I try to transfer data to USB is going to help then I will file. The only problem I see is if the maintainer or a developer asks for more info I won't be able to provide unless I see the problem again. So, what do you propose in this case? I will do accordingly. lists.opensuse.org/opensuse/2012-03/msg00702.html Best
Cheers,
Lars
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Mon, Mar 19, 2012 at 12:05:07AM +0100, Swapnil Bhartiya wrote: [ 8< ]
In my case as I said that the problem went away after reboot.
If this issie is completely gone then a bugreport doesn't make much sense. Well, you can try is to provoke it by a plug, unplug session.
I had not rebooted the system for days.
If it's more likely to trigger the issue by rebooting, then be this nice to reboot your system. I know we all don't like to reboot and we like to keep our system running. But in this case our common goal must be to get rid of this issue. And it looks like you're currently the only one willing and able to push it further.
I said I would file the report when the problem recurs. Not that I don't want to file it (it's only going to help users like me) but I don't know what to file unless I get the problem again. If just creating a bug report that my system freezes when I try to transfer data to USB is going to help then I will file. The only problem I see is if the maintainer or a developer asks for more info I won't be able to provide unless I see the problem again. So, what do you propose in this case? I will do accordingly.
We must identify a way to get into the situation as you had before again. One approach I see is the described the repeated plug and unplug of the device. The other is to reboot several times. Also try to remember if the USB device was plugged or unplugged when you ran into the issue. Thanks, Lars -- Lars Müller [ˈlaː(r)z ˈmʏlɐ] Samba Team SUSE Linux, Maxfeldstraße 5, 90409 Nürnberg, Germany
On 03/19/2012 12:28 AM, Lars Müller wrote:
On Mon, Mar 19, 2012 at 12:05:07AM +0100, Swapnil Bhartiya wrote: [ 8< ]
In my case as I said that the problem went away after reboot.
If this issie is completely gone then a bugreport doesn't make much sense.
Well, you can try is to provoke it by a plug, unplug session.
I had not rebooted the system for days.
If it's more likely to trigger the issue by rebooting, then be this nice to reboot your system. I know we all don't like to reboot and we like to keep our system running. But in this case our common goal must be to get rid of this issue. And it looks like you're currently the only one willing and able to push it further.
I said I would file the report when the problem recurs. Not that I don't want to file it (it's only going to help users like me) but I don't know what to file unless I get the problem again. If just creating a bug report that my system freezes when I try to transfer data to USB is going to help then I will file. The only problem I see is if the maintainer or a developer asks for more info I won't be able to provide unless I see the problem again. So, what do you propose in this case? I will do accordingly.
We must identify a way to get into the situation as you had before again.
One approach I see is the described the repeated plug and unplug of the device. The other is to reboot several times.
I have tried couple of times with all USBs I had. Formatted them from different machines in NTFS but no issue. :(
Also try to remember if the USB device was plugged or unplugged when you ran into the issue.
The USB device was plugged. The system would freeze after copying around 10-25% of data. If I unplugged the device while system was frozen everything will go back to normal and system will resume. Thanks for your time :-) Swapnil -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Mon, 2012-03-19 at 00:28 +0100, Lars Müller wrote:
On Mon, Mar 19, 2012 at 12:05:07AM +0100, Swapnil Bhartiya wrote: [ 8< ]
In my case as I said that the problem went away after reboot.
If this issie is completely gone then a bugreport doesn't make much sense.
Well, you can try is to provoke it by a plug, unplug session.
I had not rebooted the system for days.
If it's more likely to trigger the issue by rebooting, then be this nice to reboot your system. I know we all don't like to reboot and we like to keep our system running. But in this case our common goal must be to get rid of this issue. And it looks like you're currently the only one willing and able to push it further.
Excuse me for jumping in the middle of a thread. But i had something simular. I used to do big transfers with rsync from an USB-disk. During that time (several hours) the system, a DL380G5 was not responsive any more. Situation went back to normal after the transfer. To avoid this (50 virtual machines were running on it), i now do the transfer since then, via a NFS-connected machine. No real solution, but it prevents a storm of IRQ's hw -- 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 2012-03-18 23:45, Lars Müller wrote:
On Sun, Mar 18, 2012 at 10:23:27PM +0100, Carlos E. R. wrote:
Look, if someone doesn't see the need to communicate a bug issue the predefined way, then this issue might be less pressing. Also not all Open Source Software developers eager to be subscribed to yet another mailing list. Ask yourself why should a Linux kernel or a glibc guy hang around here?
Don't try to convince me :-) There is no need. I'm trying to help as a kind of midleware or go between. I tried to convince him and failed, and for that, I'm sorry. However, I suggest you look at his post and learn his reasons for not doing it instead of trying to guess why he does not want to do it. Maybe someone else can convince him. - -- Cheers / Saludos, Carlos E. R. (from 11.4 x86_64 "Celadon" at Telcontar) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iEYEARECAAYFAk9ma8UACgkQIvFNjefEBxqVSQCeOvSq5laC6Gm0ZoLF0RFzJlOd Dj0An1K1VQjNwuxzy2DLXNysgacCcrTZ =XWJT -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (6)
-
Carlos E. R. -
Carlos E. R. -
Graham Anderson -
Hans Witvliet -
Lars Müller -
Swapnil Bhartiya