Re: 4.4-RELEASE corrupting filesystems?
Cy Schubert - ITSD Open Systems Group wrote: In message [EMAIL PROTECTED], Vivek Khera writes: n == nsayer [EMAIL PROTECTED] writes: n One of my machines has been prone to having one of its filesystems curdle n for an unknown reason. n I have ever only experienced 3 filesystems get curdled on FreeBSD ever, n but all three were /usr filesystems on that one disk. The first time n was a while ago, but now it's happened twice in the last week. This smells *seriously* of hardware failure. Could be bad controller, disk, cable, or even memory. Only detailed diagnostics can pinpoint the real culprit (either that or sequentially replacing every part one at a a time.) Or possibly a termination problem (which is also a h/w problem). Or the NFS corruption problem that has been discussed on -hackers. Fixes for it are still developed in current. Kent -- Kent Stewart Richland, WA mailto:[EMAIL PROTECTED] http://users.owt.com/kstewart/index.html FreeBSD News http://daily.daemonnews.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: Userland not installing
Something is rather odd here, and I'm at a loss to explain it. Checking dates does in fact show a Dec 17th datestamp, but comparing the /usr/bin executables against /usr/obj/usr/src/usr.bin executables shows they are different in size, but datestamps are the same, Dec 17. This whole thing started with a funny netstat output. Looking at file sizes, I see in /usr/bin, netstat is: -r-xr-sr-x 1 root kmem 91008 Dec 17 13:06 /usr/bin/netstat and in /usr/obj/usr/src/usr.bin/netstat: -rwxr-xr-x 1 root wheel 99182 Dec 17 12:24 netstat Knowing it has not been installed, but it does give proper output. A view of the script of the install does not show any problems... More than baffled now... Sam To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: Userland not installing
Your issue is basically this? pedicular~;ls -lai /usr/obj/usr/src/usr.bin/xargs/ total 26 309805 drwxr-xr-x2 root wheel 512 Dec 12 12:24 . 238112 drwxr-xr-x 211 root wheel 6656 Dec 12 11:19 .. 310095 -rw-r--r--1 root wheel 884 Dec 12 11:58 .depend 310504 -rwxr-xr-x1 root wheel 8425 Dec 12 12:24 xargs 310505 -rw-r--r--1 root wheel 2769 Dec 12 12:24 xargs.1.gz 310503 -rw-r--r--1 root wheel 4544 Dec 12 12:24 xargs.o pedicular~;ls -lai /usr/bin/xargs 8627 -r-xr-xr-x 1 root wheel 6324 Dec 12 12:53 /usr/bin/xargs I *believe* that the install strips the binaries before installing them. That would explain the size difference. On Mon, Dec 17, 2001 at 01:45:13PM -0500, Sam Drinkard wrote: Something is rather odd here, and I'm at a loss to explain it. Checking dates does in fact show a Dec 17th datestamp, but comparing the /usr/bin executables against /usr/obj/usr/src/usr.bin executables shows they are different in size, but datestamps are the same, Dec 17. This whole thing started with a funny netstat output. Looking at file sizes, I see in /usr/bin, netstat is: -r-xr-sr-x 1 root kmem 91008 Dec 17 13:06 /usr/bin/netstat and in /usr/obj/usr/src/usr.bin/netstat: -rwxr-xr-x 1 root wheel 99182 Dec 17 12:24 netstat Knowing it has not been installed, but it does give proper output. A view of the script of the install does not show any problems... More than baffled now... Sam -- Michael Lucas [EMAIL PROTECTED], [EMAIL PROTECTED] my FreeBSD column: http://www.oreillynet.com/pub/q/Big_Scary_Daemons http://www.blackhelicopters.org/~mwlucas/ To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: Userland not installing
Date: Mon, 17 Dec 2001 13:45:13 -0500 From: Sam Drinkard [EMAIL PROTECTED] Something is rather odd here, and I'm at a loss to explain it. Checking dates does in fact show a Dec 17th datestamp, but comparing the /usr/bin executables against /usr/obj/usr/src/usr.bin executables shows they are different in size, but datestamps are the same, Dec 17. This whole thing started with a funny netstat output. Looking at file sizes, I see in /usr/bin, netstat is: -r-xr-sr-x 1 root kmem 91008 Dec 17 13:06 /usr/bin/netstat and in /usr/obj/usr/src/usr.bin/netstat: -rwxr-xr-x 1 root wheel 99182 Dec 17 12:24 netstat Knowing it has not been installed, but it does give proper output. A view of the script of the install does not show any problems... FYI: m133[1] file /usr/bin/netstat /usr/obj/usr/src/usr.bin/netstat/netstat /usr/bin/netstat: setgid ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), dynamically linked (uses shared libs), stripped /usr/obj/usr/src/usr.bin/netstat/netstat: ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), dynamically linked (uses shared libs), not stripped m133[2] ls -l !:* ls -l /usr/bin/netstat /usr/obj/usr/src/usr.bin/netstat/netstat -r-xr-sr-x 1 root kmem 91008 Dec 17 05:58 /usr/bin/netstat -rwxrwxr-x 1 root wheel 99182 Dec 17 05:38 /usr/obj/usr/src/usr.bin/netstat/netstat m133[3] (I.e., the new one *is* installed; it's stripped, while the one in /usr/obj is not stripped, thus accounting for the size difference.) Cheers, david -- David H. Wolfskill [EMAIL PROTECTED] I believe it would be irresponsible (and thus, unethical) for me to advise, recommend, or support the use of any product that is or depends on any Microsoft product for any purpose other than personal amusement. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: Userland not installing
That is true, however I can't explain why netstat -a only gives the active UNIX domain sockets info, while the unstripped version reports Active Internet connections (including servers) as it's supposed to.. Sam To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: Userland not installing
Of course, one could always use file on each binary to see what the differences are. But that would be too easy. Sigh... the new tea I tried must not have enough caffiene. On Mon, Dec 17, 2001 at 10:53:40AM -0800, David Wolfskill wrote: Date: Mon, 17 Dec 2001 13:45:13 -0500 From: Sam Drinkard [EMAIL PROTECTED] Something is rather odd here, and I'm at a loss to explain it. Checking dates does in fact show a Dec 17th datestamp, but comparing the /usr/bin executables against /usr/obj/usr/src/usr.bin executables shows they are different in size, but datestamps are the same, Dec 17. This whole thing started with a funny netstat output. Looking at file sizes, I see in /usr/bin, netstat is: -r-xr-sr-x 1 root kmem 91008 Dec 17 13:06 /usr/bin/netstat and in /usr/obj/usr/src/usr.bin/netstat: -rwxr-xr-x 1 root wheel 99182 Dec 17 12:24 netstat Knowing it has not been installed, but it does give proper output. A view of the script of the install does not show any problems... FYI: m133[1] file /usr/bin/netstat /usr/obj/usr/src/usr.bin/netstat/netstat /usr/bin/netstat: setgid ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), dynamically linked (uses shared libs), stripped /usr/obj/usr/src/usr.bin/netstat/netstat: ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), dynamically linked (uses shared libs), not stripped m133[2] ls -l !:* ls -l /usr/bin/netstat /usr/obj/usr/src/usr.bin/netstat/netstat -r-xr-sr-x 1 root kmem 91008 Dec 17 05:58 /usr/bin/netstat -rwxrwxr-x 1 root wheel 99182 Dec 17 05:38 /usr/obj/usr/src/usr.bin/netstat/netstat m133[3] (I.e., the new one *is* installed; it's stripped, while the one in /usr/obj is not stripped, thus accounting for the size difference.) Cheers, david -- David H. Wolfskill[EMAIL PROTECTED] I believe it would be irresponsible (and thus, unethical) for me to advise, recommend, or support the use of any product that is or depends on any Microsoft product for any purpose other than personal amusement. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message -- Michael Lucas [EMAIL PROTECTED], [EMAIL PROTECTED] my FreeBSD column: http://www.oreillynet.com/pub/q/Big_Scary_Daemons http://www.blackhelicopters.org/~mwlucas/ To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
on/off NFS connection errors
For a while I have been treating this as an annoyance, but I thought it would be wise to investigate if something larger and more nefarious might be being indicated by this. I have a mixed environment of mainly Linux boxen, with a dozen or so FreeBSD machines (For when we need the kind of network resources that raising the NMBCLUSTERS can offer.) Both types of systems serve mainly as webservers, serving content that ultimately comes off of exported NFS directories, from a Network Appliance (NetApp Release 5.3.4R3: Thu Jan 27 12:08:07 PST 2000) The Linux boxen don't complain at all, but the FreeBSD boxen can get rather noisy about NFS connection errors. It happens on and off like so: 118Dec 15 21:01:47 cc117 /kernel: nfs server netapp1:/vol/members: not responding 118Dec 15 21:01:47 cc117 /kernel: nfs server netapp1:/vol/members: is alive again 6nfs server netapp1:/vol/members: not responding 6nfs server netapp1:/vol/members: is alive again 118Dec 15 22:34:19 cc117 /kernel: nfs server netapp1:/vol/members: not responding 118Dec 15 22:34:19 cc117 /kernel: nfs server netapp1:/vol/members: is alive again 6nfs server netapp1:/vol/members: not responding 6nfs server netapp1:/vol/members: is alive again 118Dec 15 22:39:19 cc117 /kernel: nfs server netapp1:/vol/members: not responding 118Dec 15 22:39:19 cc117 /kernel: nfs server netapp1:/vol/members: is alive again 6nfs server netapp1:/vol/members: not responding 6nfs server netapp1:/vol/members: is alive again 118Dec 15 22:40:19 cc117 /kernel: nfs server netapp1:/vol/members: not responding 118Dec 15 22:40:19 cc117 /kernel: nfs server netapp1:/vol/members: is alive again One moment we are connected, another we are down, and the we are back up again. Some days I can get pages and pages of this, others very little. Luckily the connection error is so short lived that Apache never hiccups. Has anyone else seen these kinds of persistent NFS errors is the 4.x branch? (This didn't happen noticeably in 3.x, but I would still maintain that the NFS code in 4.x is an improvement over 3.x.) Can anyone suggest a sysctl/kernel variable I might tune to help remedy the problem? If the root of the problem is more likely on the Netapp side, I have a support contact and am not afraid to use it. Anyone have any advice or suggestions to offer? This is the platform that I am working on: FreeBSD cc117 4.2-STABLE FreeBSD 4.2-STABLE #0: Sat Aug 18 00:21:16 EDT 2001 root@cc117:/usr/src/sys/compile/CCI_KERNEL i386 - Marius M. Rex Hardware: n. The parts of a computer that can be kicked. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: 4.4-STABLE crashes - suspects new ata-driver over wd-drivers
On Fri, Dec 07, 2001 at 01:33:14PM -0800, Brady Montz wrote: Richard Nyberg [EMAIL PROTECTED] writes: On Fri, Dec 07, 2001 at 10:59:13AM +0100, Samuel Tardieu wrote: I am experiencing the same crashes on my new machine (ATA100 IDE drive): they appeared when I noticed that I had forgotten to use soft-updates. After I have turned them on, I experienced the first crash in 15 minutes. Then I get one every two days, when doing heavy disk IOs. I got a crash 10 minutes ago when the machine was unattended though (and not doing important disk IOs), and could see a panic message on the console. Unfortunately, I hadn't enough free space in /var/crash to save the kernel. Do you people use soft-updates? From my experience on this problem, I assume that either soft-updates or the ATA driver may be causing those spontanous reboots. Yes I use soft-updates. The peculiar thing aboout my crash though is that there was no panic; the machine just froze and the screen went blank, so maybe I was hit by a different problem. -Richard Yeah, I'm using soft updates too. My crashes are generally the same as Richards - no panic, just a freeze. Except my screen doesn't go blank. Here's an update ... I'm fairly certain there's a kernel bug at work here. Last night I rebooted to linux (which is on the same disk), and ran batch compiles all night long without any troubles. In comparision, I can't compile more than an hour at a time with BSD 4.4 before it crashes. Again, I ran memtest86 and it didn't find any memory errors, and I'm not seeing any file system corruption, just hangs and reboots. I am running the latest 4.4-stable. The other day I went back to 4.4-release and that didn't help. I've tried both with and without softupdates. The crashes seem to happen most often when accessing stuff from all over the filesystem, such as during a large make clean, or most reliably, with portsdb -Uu. I am tiring of this. Someone else on this thread mentioned that their 5.0 machine is doing fine. In what shape is that and how much effort is it to move a 4.4 machine to 5.0? -- Brady Montz [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: on/off NFS connection errors
In message [EMAIL PROTECTED], Marius M. Rex writes: 118Dec 15 22:40:19 cc117 /kernel: nfs server netapp1:/vol/members: not responding 118Dec 15 22:40:19 cc117 /kernel: nfs server netapp1:/vol/members: is alive again ... Has anyone else seen these kinds of persistent NFS errors is the 4.x branch? These are a side-effect of the operation of the NFS dynamic retransmit timeout code. The NFS client measures the request response time for various types of operations and it sets a timeout based on the mean and deviation of observed times. The time taken by the server to perform some operations can vary wildly though, so occasionally when a large number of operations complete with very little delay, the response time estimate and hence the timeout become very small. Then when one request is unusually slow to complete (such as when the disk on the server is busy), the client thinks that the server isn't responding and prints those warnings. A fraction of a second later the request completes and the client prints a an 'is alive again' message. On non-soft mounts these messages are completely harmless because the client will just wait for the server to eventually reply. On soft mounts, the effect can cause problems because applications occasionally see an EINTR error. The dynamic retransmit timeout code can be disabled with the `-d' flag to mount_nfs; this is often recommended for fast networks that see very little packet loss. Ian To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: on/off NFS connection errors
:For a while I have been treating this as an annoyance, but I thought it :would be wise to investigate if something larger and more nefarious might :be being indicated by this. : :I have a mixed environment of mainly Linux boxen, with a dozen or so :FreeBSD machines (For when we need the kind of network resources that :raising the NMBCLUSTERS can offer.) Both types of systems serve mainly as :webservers, serving content that ultimately comes off of exported NFS :directories, from a Network Appliance (NetApp Release 5.3.4R3: Thu Jan 27 :12:08:07 PST 2000) The Linux boxen don't complain at all, but the FreeBSD :boxen can get rather noisy about NFS connection errors. It happens :on and off like so: : :118Dec 15 21:01:47 cc117 /kernel: nfs server netapp1:/vol/members: not :responding :118Dec 15 21:01:47 cc117 /kernel: nfs server netapp1:/vol/members: is :alive again :.. Our NFS is somewhat finicky about response times. It could probably use some tuning. Trying mounting with the 'dumbtimer' mount option and see if that fixes your problem. (also see the -d and -t options in 'man mount_nfs'. Note that -t is specified in 1/10 second increments). -Matt :Has anyone else seen these kinds of persistent NFS errors is the 4.x :branch? (This didn't happen noticeably in 3.x, but I would still :maintain that the NFS code in 4.x is an improvement over 3.x.) Can anyone :suggest a sysctl/kernel variable I might tune to help remedy the problem? :If the root of the problem is more likely on the Netapp side, I have a support :contact and am not afraid to use it. Anyone have any advice or :suggestions to offer? : :This is the platform that I am working on: :FreeBSD cc117 4.2-STABLE FreeBSD 4.2-STABLE #0: Sat Aug 18 00:21:16 EDT :2001 root@cc117:/usr/src/sys/compile/CCI_KERNEL i386 To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: 4.4-STABLE crashes - suspects new ata-driver over wd-drivers
On Mon, Dec 17, 2001 at 04:14:08PM -0800, Brady Montz wrote: Here's an update ... I'm fairly certain there's a kernel bug at work here. Last night I rebooted to linux (which is on the same disk), and ran batch compiles all night long without any troubles. In comparision, I can't compile more than an hour at a time with BSD 4.4 before it crashes. I still agree. My -current machines run find, and I refuse to run -stable on an SMP machine. I am running the latest 4.4-stable. The other day I went back to 4.4-release and that didn't help. I've tried both with and without softupdates. The crashes seem to happen most often when accessing stuff from all over the filesystem, such as during a large make clean, or most reliably, with portsdb -Uu. I am tiring of this. Someone else on this thread mentioned that their 5.0 machine is doing fine. In what shape is that and how much effort is it to move a 4.4 machine to 5.0? Unless you feel confident that you can deal with the problems that arise on -current, I wouldn't want to be the one to recommend that you change, but my personal experience has been that -stable is anything but stable on SMP machines. On UP machines I have no problems at all. The -current SMP machines here are all very stable. I don't track it daily, and I am careful to build a test box before I rebuild a box I care about, but generally I have been much happier with -current than with -stable (this year). As for the effort to upgrade, it depends on what the box is doing. I've only upgraded a few boxes in the last year or so, and found that it was fairly timing dependant, but in general I haven't had any real problems (read UPDATING). I have a little time this afternoon, so I'm going to see if I can figure something out. I'll throw -stable onto one of my SMP development machines and see if I can kill it. At least there I can debug it. A small plug: I've written a script that will rebuild an entire machine, from a cvsup - mergemaster and reboot. It doesn't really address anything to do with this thread, but you might find it handy :) http://www.acns.ab.ca/projects/rebuild/rebuild.tar.gz -- Chad David[EMAIL PROTECTED] ACNS Inc. Calgary, Alberta Canada To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: 4.4-STABLE crashes - suspects new ata-driver over wd-drivers
Date: Mon, 17 Dec 2001 14:21:15 -0700 From: Chad David [EMAIL PROTECTED] I still agree. My -current machines run find, and I refuse to run -stable on an SMP machine. For whatever it may be worth, my build machine (which is one of the machines on which I track both -STABLE and -CURRENT daily) is an SMP box. It also has a local CVS repository on it (from which I update the CVS repository on my laptop, which laso tracks -STABLE and -CURRENT daily). I am not having any problems with -STABLE that I know of. Cheers, david -- David H. Wolfskill [EMAIL PROTECTED] I believe it would be irresponsible (and thus, unethical) for me to advise, recommend, or support the use of any product that is or depends on any Microsoft product for any purpose other than personal amusement. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: staroffice5.2
Hello Antonin I've tested also serveral time but with no success. Afterwoods I used the original SUN cd-rom. Do you have one? On Mon, Dec 17, 2001 at 04:35:46PM +0100 Ing. Antonín Walter wrote: maybe I did not use the correct word in english, but on using the port you have suggested it first downloaded the 97GB file of staroffice and the tried to download this 109939-02.tar.Z file but it was not available at any location the port was looking at -- Regards Martin Schweizer [EMAIL PROTECTED] PC-Service M. Schweizer; Gewerbehaus Schwarz; CH-8608 Bubikon Tel. +41 55 243 30 00; Fax: +41 55 243 33 22; http://www.pc-service.ch To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: 4.4-STABLE crashes - suspects new ata-driver over wd-drivers
On Mon, Dec 17, 2001 at 01:33:19PM -0800, David Wolfskill wrote: Date: Mon, 17 Dec 2001 14:21:15 -0700 From: Chad David [EMAIL PROTECTED] I still agree. My -current machines run find, and I refuse to run -stable on an SMP machine. For whatever it may be worth, my build machine (which is one of the machines on which I track both -STABLE and -CURRENT daily) is an SMP box. It also has a local CVS repository on it (from which I update the CVS repository on my laptop, which laso tracks -STABLE and -CURRENT daily). I am not having any problems with -STABLE that I know of. It might be worth a lot :). It is possible that I and a few others have bad hardware, and there is no problem with -stable. It seems unlikely, but it is not at all impossible I guess. Are you running any combination of samba/nfs/ata? Thanks -- Chad David[EMAIL PROTECTED] ACNS Inc. Calgary, Alberta Canada To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
