Re: [zfs-discuss] (Fletcher+Verification) versus (Sha256+No Verification)
Edward, this is OT but may I suggest you to use something like Wolfram Alpha to perform your calculations a bit more comfortably? -- Enrico M. Crisostomo On Jan 12, 2011, at 4:24, Edward Ned Harvey opensolarisisdeadlongliveopensola...@nedharvey.com wrote: For anyone who still cares: I'm calculating the odds of a sha256 collision in an extremely large zpool, containing 2^35 blocks of data, and no repetitions. The formula on wikipedia for the birthday problem is: p(n;d) ~= 1-( (d-1)/d )^( 0.5*n*(n-1) ) In this case, n=2^35 d=2^256 The problem is, this formula does not compute because n is so large. Fortunately x = e^ln(x) and so we're able to use this technique to make the huge exponent instead, a huge multiplication. (Using the bc mathlib notation, the l(x) function is the natural log of x, and e(x) is e raised to the power of x) p(n;d) ~= 1-e( ( 0.5*n*(n-1)*l((d-1)/d) ) ) Using bc to calculate the answer: bc -l n=2^35 d=2^256 scale=1024 1-e( ( 0.5*n*(n-1)*l((d-1)/d) ) ) .50978941154 I manually truncated here (precision goes out 1024 places). This is 5.1*10^-57 Note: I had to repeat the calculation many times, setting a larger and larger scale. The default scale of 20, and even 64 and 70 and 80 were not precise enough to produce a convergent answer around the -57th decimal place. So I just kept going larger, and in retrospect, anything over 100 would have been fine. I wrote 1024 above, so who cares. If you've been paying close attention you'll recognize this is the same answer I originally calculated, but my original equation was in fact wrong. It just so happens that my original equation neglects the probability of collisions from previous events, so it is actually accurate whenever the probability of previous events is insignificant. It is merely luck that for the data size in question, my equation produced something that looked correct. It would produce a wrong calculation of probability for a larger value of n. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS Storage server hardwae
I'm currently using SXCE with eSATA (with an LSI controller) and SAS disks in my home boxes and they run just fine. The only glitch I had after LU-upgrading to the latest release is eSATA disk not spinning down any longer when idle. I export file systems with NFS to my Macs: beware that Mac OS X uses decomposed UTF-8 characters and sometimes I have some portability issues when file names contain, for example, accented characters. It runs fine and pretty better than CIFS, IMHO. In some case I use an OS X iSCSI initiator and Comstar: it runs fine and it's the only solution I found if you need, for example, to use time machine upon a ZFS volume. Bye, Enrico -- Enrico M. Crisostomo On Aug 25, 2010, at 21:29, Dr. Martin Mundschenk m.mundsch...@me.com wrote: Hi! I'm running a OSOL box for quite a while and I think ZFS is an amazing filesystem. As a computer I use a Apple MacMini with USB and FireWire devices attached. Unfortunately the USB and sometimes the FW devices just die, causing the whole system to stall, forcing me to do a hard reboot. I had the worst experience with an USB-SATA bridge running an Oxford chipset, in a way that the four external devices stalled randomly within a day or so. I switched to a four slot raid box, also with USB bridge, but with better reliability. Well, I wonder what are the components to build a stable system without having an enterprise solution: eSATA, USB, FireWire, FibreChannel? Martin ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Is it safe/possible to idle HD's in a ZFS Vdev to save wear/power?
Hi. I'm using two SIIG eSATA II PCIe PRO adapters on a Sun Ultra 24 workstation, too. The adapters are connected to four external eSATA drives that made up a zpool used for scheduled back-up purposes. I'm now running SXCE b129, live upgraded from b116. Before the live upgrade the external disks were spinning down when idle, now they never do. /etc/power.conf was not modified. I'm probably waiting the next OpenSolaris release and check if it works. Nevertheless, in the meanwhile, do you have any suggestion about how to debug this? I don't like wasting energy this way and I cannot shutdown that machine so often. Thanks, Enrico On Sat, Apr 17, 2010 at 8:43 AM, Bill Sommerfeld bill.sommerf...@oracle.com wrote: On 04/16/10 20:26, Joe wrote: I was just wondering if it is possible to spindown/idle/sleep hard disks that are part of a Vdev pool SAFELY? it's possible. my ultra24 desktop has this enabled by default (because it's a known desktop type). see the power.conf man page; I think you may need to add an autopm enable if the system isn't recognized as a known desktop. the disks spin down when the system is idle; there's a delay of a few seconds when they spin back up. - Bill ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] automate zpool scrub
It's ok to use zpool full path. Nontheless, I'd suggest you read crontab man page to learn how you can set some options, such as paths, shell, timezones and so on directly into your crontab files. On Nov 1, 2009, at 16:45, Vano Beridze vanua...@gmail.com wrote: Now I've logged in and there was a mail saying that cron did not found zpool it's in my path which zpool /usr/sbin/spool Does cron use different PATH setting? Is it ok to specify /usr/sbin/zpool in crontab file? -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] automate zpool scrub
Glad it helped you. As far as it concerns your observation about the root user, please take into account that Solaris Role Based Access control lets you fine tune privileges you grant to users: your ZFS administrator needs not be root. Specifically, if you have a look at your /etc/prof_attr and /etc/exec_attr, you'll notice that there exist two profiles: ZFS Storage Management and ZFS File System Management: exec_attr:ZFS File System Management:solaris:cmd:::/sbin/zfs:euid=0 exec_attr:ZFS Storage Management:solaris:cmd:::/sbin/zpool:uid=0 You can run the zfs and zpool command from a mortal user account with pfexec if such users is associated with the corresponding profile. Bye, Enrico On Sun, Nov 1, 2009 at 9:03 PM, Vano Beridze vanua...@gmail.com wrote: I've looked at man cron and found out that I can modify /etc/default/cron file to set PATH that is defaulted for /usr/bin for mortal users and /usr/bin:/usr/sbin for root. I did not change /etc/default/cron file, instead I've indicated full path in my crontab file. Ethically speaking I guess scrubbing filesystem weekly is an administrative task and it's more applicable to root user, So If I had created crontab job for root user the whole PATH problem would not arise. Anyways it's my desktop so I'm the man and woman in here and there is no big difference what user's crontab will do the job. :) -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] White box server for OpenSolaris
On Fri, Sep 25, 2009 at 10:56 PM, Toby Thain t...@telegraphics.com.au wrote: On 25-Sep-09, at 2:58 PM, Frank Middleton wrote: On 09/25/09 11:08 AM, Travis Tabbal wrote: ... haven't heard if it's a known bug or if it will be fixed in the next version... Out of courtesy to our host, Sun makes some quite competitive X86 hardware. I have absolutely no idea how difficult it is to buy Sun machines retail, Not very difficult. And there is try and buy. Indeed, at least in Spain and in Italy I had no problem buying workstations. Recently I owned both Sun Ultra 20 M2 and Ultra 24. I had a great feeling with them and price seemed very competitive to me, compared to offers of other mainstream hardware providers. People overestimate the cost of Sun, and underestimate the real value of fully integrated. +1. People like fully integration when it comes, for example, to Apple, iPods and iPhones. When it comes, just to make another example..., to Solaris, ZFS, ECC memory and so forth (do you remember those posts some time ago?), they quickly forget. --Toby but it seems they might be missing out on an interesting market - robust and scalable SOHO servers for the DYI gang ... Cheers -- Frank ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] libzfs.h versioning
Richard, I compared the libzfs_jni source code and they're pretty different from what we're doing. libzfs_jni is essentially a jni wrapper to (yet?) another set of zfs-related programs written in C. zfs for Java, on the other hand, is a Java wrapper to the functionality of (and only of) libzfs. I suppose that libzfs_jni capabilities could be implemented on top of zfs for java but the approach is pretty different: the main difference is the purpose of the exposed methods: libzfs is the interface to ZFS and its methods are low level while libzfs_jni exposes a set of operations which are coarse grained and targeted to management. Nevertheless, the functionality provided by libzfs_jni is interesting and I'd like to build something similar by using zfs for java. Personally, I'm doing this for two reasons: having a libzfs wrapper for Java seems like a good thing to have and I'd like to use to build some management interfaces (such as web but not only) instead on having to rely on shell scripting with zfs and zpool commands. I'll keep an eye to libzfs_jni. Now, to return to the original question, I haven't found a way to correlate libzfs.h versions (and dependencies) to Nevada releases. At the moment, I'm willing to extract information from a sysinfo call (any suggestion about a better way?) and the next step, whose logic I'm missing, is how to correlate this information with to a concrete libzfs.h version from openGrok: maybe it's just trivial, but I do not find it. Have you got some information to help me address this problem? Thanks, Enrico On Fri, Sep 11, 2009 at 12:53 AM, Enrico Maria Crisostomo enrico.m.crisost...@gmail.com wrote: On Fri, Sep 11, 2009 at 12:26 AM, Richard Elling richard.ell...@gmail.com wrote: On Sep 10, 2009, at 1:03 PM, Peter Tribble wrote: On Thu, Sep 10, 2009 at 8:52 PM, Richard Elling richard.ell...@gmail.com wrote: Enrico, Could you compare and contrast your effort with the existing libzfs_jni? http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/libzfs_jni/common/ Where's the source for the java code that uses that library? Excellent question! It is used for the BUI ZFS manager webconsole that comes with S10 and SXCE. So you might find the zfs.jar as /usr/share/webconsole/webapps/zfs/WEB-INF/lib/zfs.jar The jar only contains the class files, though. Yes, that's what I thought when I saw it. Furthermore, the last time I tried it was still unaligned with the new ZFS capabilites: it crashed because of an unknown gzip compression type... Someone from Sun could comment on the probability that they will finally get their act together and have a real BUI framework for systems management... they've tried dozens (perhaps hundreds) of times, with little to show for the effort :-( By the way, one of the goals I'd like to reach with such kind of library is just that: putting the basis for building a java based management framework for ZFS. Unfortunately wrapping libzfs will hardly fulfill this goal and the more I dig into the code the more I realize that we will need to wrap (or reimplement) some of the logic of the zfs and zpool commands. I'm also confident that building a good library on top of this wrapper will give us a very powerful tool to play with from Java. -- richard -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] libzfs.h versioning
Hi. I'm willing to maintain a project hosted on java.net (https://zfs.dev.java.net/) that aims to provide a Java wrapper to libzfs. I've already wrapped, although not committed yet, the last libzfs.h I found on OpenSolaris.org (v. 10342:108f0058f837) and the first problem I want to address is library versioning. The existing sources are wrapping an old version of libzfs.h and, as far as I can see, there were changes in libzfs.h history which would disrupt the wrapper functionality and I just wouldn't like to present the user with linker errors. Rather, I'd like to keep track of libzfs.h history during the various Nevada builds and plug the correct wrapper at runtime, during the library bootstrap. Obviously, an user could just choose and use directly the wrapper, which will be the equivalent of linking against libzfs, and do what it wants to with it. Our idea, which the project founder has already brought on, is wrapping much of the functionality (if not all...) behind a hierarchy of Java classes which would take care of the implementation details and shield the user against library changes. The first question is, then: how can I determine which libzfs.h version has gone in which Nevada build? Once I have this information, how would you suggest me to plug the wrapper at runtime? I was thinking about something like uname -rv and use that information to load wrappers, but perhaps there are finest ways to do this. Thanks in advance, Enrico -- Enrico M. Crisostomo ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] libzfs.h versioning
Thanks for pointing it out, Richard. I missed libzfs_jni. I'll have a look at it and see where we're overlapping. As far as I can see at a quick glance is that libzfs_jni is including functionality we'd like to build upon the libzfs wrapper (that's why I was studying zfs and zpool commands). Maybe a convergence may be worthwhile: I'll study it ASAP. Thanks for the pointer Richard! Enrico The first thing On Thu, Sep 10, 2009 at 9:52 PM, Richard Elling richard.ell...@gmail.com wrote: Enrico, Could you compare and contrast your effort with the existing libzfs_jni? http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/libzfs_jni/common/ Perhaps it would be worthwhile to try and un-privatize libzfs_jni? -- richard On Sep 10, 2009, at 12:20 PM, Enrico Maria Crisostomo wrote: Hi. I'm willing to maintain a project hosted on java.net (https://zfs.dev.java.net/) that aims to provide a Java wrapper to libzfs. I've already wrapped, although not committed yet, the last libzfs.h I found on OpenSolaris.org (v. 10342:108f0058f837) and the first problem I want to address is library versioning. The existing sources are wrapping an old version of libzfs.h and, as far as I can see, there were changes in libzfs.h history which would disrupt the wrapper functionality and I just wouldn't like to present the user with linker errors. Rather, I'd like to keep track of libzfs.h history during the various Nevada builds and plug the correct wrapper at runtime, during the library bootstrap. Obviously, an user could just choose and use directly the wrapper, which will be the equivalent of linking against libzfs, and do what it wants to with it. Our idea, which the project founder has already brought on, is wrapping much of the functionality (if not all...) behind a hierarchy of Java classes which would take care of the implementation details and shield the user against library changes. The first question is, then: how can I determine which libzfs.h version has gone in which Nevada build? Once I have this information, how would you suggest me to plug the wrapper at runtime? I was thinking about something like uname -rv and use that information to load wrappers, but perhaps there are finest ways to do this. Thanks in advance, Enrico -- Enrico M. Crisostomo ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] libzfs.h versioning
On Fri, Sep 11, 2009 at 12:26 AM, Richard Elling richard.ell...@gmail.com wrote: On Sep 10, 2009, at 1:03 PM, Peter Tribble wrote: On Thu, Sep 10, 2009 at 8:52 PM, Richard Elling richard.ell...@gmail.com wrote: Enrico, Could you compare and contrast your effort with the existing libzfs_jni? http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/libzfs_jni/common/ Where's the source for the java code that uses that library? Excellent question! It is used for the BUI ZFS manager webconsole that comes with S10 and SXCE. So you might find the zfs.jar as /usr/share/webconsole/webapps/zfs/WEB-INF/lib/zfs.jar The jar only contains the class files, though. Yes, that's what I thought when I saw it. Furthermore, the last time I tried it was still unaligned with the new ZFS capabilites: it crashed because of an unknown gzip compression type... Someone from Sun could comment on the probability that they will finally get their act together and have a real BUI framework for systems management... they've tried dozens (perhaps hundreds) of times, with little to show for the effort :-( By the way, one of the goals I'd like to reach with such kind of library is just that: putting the basis for building a java based management framework for ZFS. Unfortunately wrapping libzfs will hardly fulfill this goal and the more I dig into the code the more I realize that we will need to wrap (or reimplement) some of the logic of the zfs and zpool commands. I'm also confident that building a good library on top of this wrapper will give us a very powerful tool to play with from Java. -- richard -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Failure of Quicktime *.mov files after move to zfs disk
Hi Harry. Glad to hear you solved the problem. As soon as I saw that Quicktime's error I thought about permissions. Unfortunately I sort of got used to it. To follow on this discussion, I think there's something strange here. It might simply be a Quicktime idiosyncrasy, which is chocking onto those permissions because it's expecting some more. Although it's a CIFS-related discussion rather than ZFS, I want to make my point clear. My reasoning is: I start with what I think is a reasonable (and intuitive) choice: mode 600 for my files (no other ACL). If I examine permissions on the Windows side they are pretty odd: as I said, there's no read there. If you examine special permissions, my user is indeed allowed List folders/Read data (amongst others). Nevertheless, almost everything works as expected. I can indeed read the files' contents and copy them locally, although I cannot launch files with Quicktime directly on the share. That's why I think Quicktime is chocking somewhere. If I give my user the Windows' read permission, on the Solaris side it materializes like this: $ ls -dV IMG_0003.JPG --+ 1 enrico staff1082821 Jul 17 19:39 IMG_0003.JPG everyone@:rwxp---A-W-Co-:---:deny user:enrico:--x---:---:deny group:staff:rwxp--:---:deny everyone@:--a-R-c--s:---:allow user:enrico:rw-p--aARWc--s:---:allow user:enrico:rw-p---A-W-Cos:---:allow group:staff:-s:---:allow I'd like to find some documentation about how we should map Windows' permissions to Solaris ACLs. It doesn't seem so intuitive to me, so far. By the way: when I set up CIFS shares and then manage them from Windows, I don't usually encounter any problem. The first time I saw the strange Quicktime behaviour was due to a cron-scheduled script which was resetting permissions on the shared files to Solaris friendly values such as 600 or 644 for files and 700 and 755 for directories. I thought it was sufficient, after mapping users and groups with idmap, but it seems to fall short sometimes. Thank you for any pointer, Enrico On Sat, Aug 22, 2009 at 9:00 PM, Harry Putnamrea...@newsguy.com wrote: Scott Laird sc...@sigkill.org writes: Checksum all of the files using something like md5sum and see if they're actually identical. Then test each step of the copy and see which one is corrupting your files. On Fri, Aug 21, 2009 at 1:43 PM, Harry Putnamrea...@newsguy.com wrote: [...] I didn't do that since I've found that opening the file from vista with a file browser started as `administator' worked.. so apparently the files are indenticle enough to play in quicktime player started as administrator . Enrico Maria Crisostomo enrico.m.crisost...@gmail.com writes: [...] Thanks for the input... I've found now that the directory on zfs server that I was scping the files to had not gotten included in a previous chmod cmd run on the zfs server. chmod -R A=everyone@:full_set:fd:allow That particular directory where the transferred files were landing, was created after having run the chmod cmd above on the server. Here something's missing to me and documentation hasn't helped me (yet)... There's no read here. Just set it (on the Vista side it's just one click) and Quicktime will work. I've got a script which resets my files' permissions something like: find /yourdir -type f -exec chmod A- {} + find /yourdir -type f -exec chmod 644 {} + The chmod command I mentioned above appears many times on the cifs-server list, as a way to avoid permissions problems and as it turns out it works in my case too. ... although it appears to only be of use when called after the files are transferred. The commands you show also appear to make things work... however, at first I thought the executable bits that are set when the files are created on windows... doesn't really seem to prevent running it from a remote vista laptop. It appears here that the permission bits as they occur work, and so does chmoding to 644. [...] Chris Wrote: It might be worth checking if they've got funny Unicode chars in the names. What normalization's happening on both servers, what version of NFS is being used? How big are the files? Apparently not the problem in my case... thanks for the input. Thomas Burgess wonsl...@gmail.com writes: i had something similar happen to me when i switched to ZFS but it turned out to be an error with cpio and the mkv format...i'm not sure exactly why but whenever i tried to backup mkv files with cpio onto ZFS it would give me corrupted files. This was also apparently not the problem in my case... thanks for the input. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Ελευθερία ή θάνατος Programming today is a race between software
Re: [zfs-discuss] Failure of Quicktime *.mov files after move to zfs disk
Check with Vista if you have permissions to read the file. I experienced the same problem (that's why I posted another questions to the CIFS mailing list about mapping users with idmap). It always happens when I copy these files from the iPhone. These files result with such permissions: $ ls -dV IMG_0004.MOV -rw--- 1 enrico staff12949182 Jul 17 19:39 IMG_0004.MOV owner@:--x---:---:deny owner@:rw-p---A-W-Co-:---:allow group@:rwxp--:---:deny group@:--:---:allow everyone@:rwxp---A-W-Co-:---:deny everyone@:--a-R-c--s:---:allow On the Vista side, having mapped with idmap even the staff group to the corresponding Windows group of the enrico user, the file results with the following special permissions for the user enrico: List folder/read data Create files/write data Create folders/append data Write attributes Write extended attributes Here something's missing to me and documentation hasn't helped me (yet)... There's no read here. Just set it (on the Vista side it's just one click) and Quicktime will work. I've got a script which resets my files' permissions something like: find /yourdir -type f -exec chmod A- {} + find /yourdir -type f -exec chmod 644 {} + Hope this helps, Enrico On Fri, Aug 21, 2009 at 11:35 PM, Scott Lairdsc...@sigkill.org wrote: Checksum all of the files using something like md5sum and see if they're actually identical. Then test each step of the copy and see which one is corrupting your files. On Fri, Aug 21, 2009 at 1:43 PM, Harry Putnamrea...@newsguy.com wrote: During the course of backup I had occassion to copy a number of quicktime video (*.mov) files to zfs server disk. Once there... navigating to them with quicktime player and opening results in a failure that (From windows Vista laptop) says: error --43: A file could not be found (Welcome.mov) I would have attributed it to some problem from scping it to the zfs server had it not been for finding that if I scp it to a linux server the problem does not occur. Both the zfs and linux (Gentoo) servers are on a home lan.. but using the same router/switch[s] over gigabit network adaptors. On both occasions the files were copied using cygwin/ssh on a Vista laptop. Anyone have an idea what might cause this. Any more details I can add that would make diagnostics easier? ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss -- Ελευθερία ή θάνατος Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning. GPG key: 1024D/FD2229AF ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] zfs send -R core dumps on SXCE 110
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Thanks Matthew, I was not sure about it. I'm looking forward to the next SXCE build. Cheers, Enrico Matthew Ahrens wrote: Enrico Maria Crisostomo wrote: # zfs send -R -I @20090329 mypool/m...@20090330 | zfs recv -F -d anotherpool/anotherfs I experienced core dumps and the error message was: internal error: Arg list too long Abort (core dumped) This is 6801979, fixed in build 111. --matt -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (SunOS) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAknWJTsACgkQW8+x8v0iKa9WIACdGkMA8ccKox8E1GkNtmvIfuJ5 3IUAnRtAm/Nb+EGavpPV0BVpJUCY2JCu =WBH8 -END PGP SIGNATURE- ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] zfs send -R core dumps on SXCE 110
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi. I'm running a live upgraded (from b104) SXCE build 110 and I'm experiencing a core dumps sending and receiving ZFS snapshots. I sent a replication stream of a snapshot with # zfs send -R mypool/m...@20090329 | zfs recv -d anotherpool/anotherfs and it went fine. Back then I didn't pay attention to the fact that @20090329 was the only snapshot existing at the moment. When I tried to send incremental replication streams with: # zfs send -R -I @20090329 mypool/m...@20090330 | zfs recv -F -d anotherpool/anotherfs I experienced core dumps and the error message was: internal error: Arg list too long Abort (core dumped) Running pstack on the core file says: $ pstack core core 'core' of 11853: zfs recv -F -d tank/backup/solaris/filesystems fef022a5 _lwp_kill (1, 6, 8042e78, feeaab7e) + 15 feeaab8a raise(6, 0, 8042ec8, fee81ffa) + 22 fee8201a abort(8042ef8, feb65000, 8042ef8, 8088670, 8088a70, 400) + f2 feaf2d59 zfs_verror (8088648, 80f, feb4d934, 8042f2c) + d5 feaf30be zfs_standard_error_fmt (8088648, 7, feb4d934, 8043fb0, 80457a0) + 1ea feaf2ec4 zfs_standard_error (8088648, 7, 8043fb0, feb0ba1d) + 28 feb0bc2b zfs_receive_one (8088648, 0, 8046e6b, a, 80457a0, 8045d00) + 15cf feb0c8e6 zfs_receive_impl (8088648, 8046e6b, a, 0, 8084788, 8046c2c) + 6b2 feb0a275 zfs_receive_package (8088648, 0, 8046e6b, a, 8046540, 8046680) + 485 feb0c8bd zfs_receive_impl (8088648, 8046e6b, a, 0, 0, 8046c2c) + 689 feb0ca49 zfs_receive (8088648, 8046e6b, a, 0, 0, 8046d00) + 35 0805847a zfs_do_receive (4, 8046d00, 8046cfc, 807187c) + 172 0805bbcf main (5, 8046cfc, 8046d14, feffb7b4) + 2af 08053f2d _start (5, 8046e5c, 8046e60, 8046e65, 8046e68, 8046e6b) + 7d I now realize that even the first command fails if I try to send a snapshot which depends on previous ones: # zfs send -R mypool/m...@20090402 | zfs receive -d anotherpool/anotherfs internal error: Arg list too long Abort (core dumped) pstack on the core file says:$ pstack core core 'core' of 5450:zfs receive -d tank/backup/solaris/filesystems fef022a5 _lwp_kill (1, 6, 80432c8, feeaab7e) + 15 feeaab8a raise(6, 0, 8043318, fee81ffa) + 22 fee8201a abort(8043348, feb65000, 8043348, 8088670, 8088a70, 400) + f2 feaf2d59 zfs_verror (8088648, 80f, feb4d934, 804337c) + d5 feaf30be zfs_standard_error_fmt (8088648, 7, feb4d934, 8044400, 8045bf0) + 1ea feaf2ec4 zfs_standard_error (8088648, 7, 8044400, feb0ba1d) + 28 feb0bc2b zfs_receive_one (8088648, 0, 80472ab, 2, 8045bf0, 8046150) + 15cf feb0c8e6 zfs_receive_impl (8088648, 80472ab, 2, 0, 8084428, 804707c) + 6b2 feb0a275 zfs_receive_package (8088648, 0, 80472ab, 2, 8046990, 8046ad0) + 485 feb0c8bd zfs_receive_impl (8088648, 80472ab, 2, 0, 0, 804707c) + 689 feb0ca49 zfs_receive (8088648, 80472ab, 2, 0, 0, 8047150) + 35 0805847a zfs_do_receive (3, 8047150, 804714c, 807187c) + 172 0805bbcf main (4, 804714c, 8047160, feffb7b4) + 2af 08053f2d _start (4, 804729c, 80472a0, 80472a8, 80472ab, 0) + 7d Any hint about diagnosing what might be going on? Thanks in advance, Enrico -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (SunOS) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAknVUEIACgkQW8+x8v0iKa8sPACfRbhkf8hY6LQFWgvQhWOyEN1/ TNAAn0QqDTTD0NfZJvy/BbEgF5MITdl/ =A2Vy -END PGP SIGNATURE- ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss