Re: [osol-discuss] pfexec does not work any longer
> >Brian, > >Thanks for the information, but it does not work. I will have another >look at the documentation, since I may be missing something in the new >release. > >Andrew pfexec was only intended for "roles" and not for ordinary users; the Primary Administrator was dangerous but when it was assigned to a role, it wasn't that dangerous as assigning it to a user. The first one requires a second password "su role cmd" but the latter one makes it easy to exploit the system from any account assigned that role. We felt that having "pfexec id" print "uid=0(root) gid=0(root)" is a bug, not a feature. There is no profiles in Solaris 11 which delivers that functionality. You can tell "su" to work like sudo by enabling pam_tty_tickets.so(1) (Yes, it is a bug that it is in that section and with that name) With the old "Primary Administrator" was really clear when you run with a profile shell: the shell is started as root and you get a "#" prompt. Of course, you can reinvent "Primary Administrator" but we recommend against that. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Is raidz root support on Solaris 11
> >Can I ask if raidz is support on root with Solaris 11? > >I am having problem doing it from auto-installer and I just thought that >may be the reason. No possible; only a mirror is allowed (and only standard compression, not gzip). Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] New problem?
It's an issue in format. Don't relabel your disks and get an older version of format. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] svccfg question
>Hi > >After trying to "svccfg import /var/svc/manifest/site/nag.xml" > >It is showing "maintenance" but not "online" > >maintenance =A0 =A0 10:18:54 svc:/site/nag:default > >What is this problem? First start reading the log file; the system didn't start properly. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] too many tcp connection in close-wait
>Hi, > > I use solaris Unix . > >I find there is some problem in application and it generate many "close-wait" >tcp connect and stay in the server . it is generate by process id 7740 > >[CODE] > >root@XX # netstat -an | grep CLOSE_WAIT | wc -l > > 285 > > > > > >root@XX # /usr/local/bin/lsof -i:34805 > >COMMANDPIDUSER FD TYPEDEVICE SIZE/OFF NODE NAME > >XXX 7740 user 87u IPv4 0x30004ce4c48 0t1500 TCP xx-1:34805->xx-2:32924 >(CLOSE_WAIT) > It's a bug in the application; it forgets to close the TCP sockets. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] firewall and tcpwrapper help
>Thank you > >inetadm -p >The tcp_wrappers=3DFalse > >How can I set it to TRUE? Any file is for this setting? You do that with inetadm. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] firewall and tcpwrapper help
>Hi > >Now I try to use inetadm but there is no tcpwrapper there. inetadm | >grep tcpwrapper > >Does it mean tcpwrapper is not in the system? No. tcpwrappers is not a service, it a property of all of the other services. >How can I install it? It is part of the base OS. >For the firewall, can I use the GUI interface to configure it? ip.conf > Start from "inetadm -p" and look at the tcp_wrappers option. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Zones' init broken after upgrade from OpenSolaris 134b to Solaris11Express
> >Nov 20 10:11:25 silence genunix: [ID 729207 kern.warning] WARNING: init(1M) >for zone router2 (pid 1326) core dumped on signal 12: restarting automatically >Nov 20 10:11:25 silence genunix: [ID 729207 kern.warning] WARNING: init(1M) >for zone buildbot (pid 1306) core dumped on signal 12: restarting automatically > >and the /var/adm/messages is quite full of those messages. >If this is a bug, is it already reported or shall I do it myself? You need to upgrade your zones. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>Looks to me like >http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/libc/port/threads/thr.c > >checks sysconf(_SC_STACK_PROT) and honors it. The only problem is >that it caches the return value (static variable stackprot), which >would have to be invalidated by the mprotect() wrapper when the magic >ADDR_STACK was used (so that it would pick up the new process-wide >value that was also set in the proc structure). That would make that >setting apply to all future thread stacks. > >In other words, that source file would have to add a _-prefixed library >internal function that zeros stackprot. > >Maybe the best one can do is the main stack (current+future) and >the thread stacks (future), if one can't readily figure out from >either user space (syscall wrapper check) or kernel space the location >and size of all the current thread stacks. > >I haven't looked at how sigaltstack() handling might get involved... It's created by the application and so it's either rwx (32 bit or sparc) or rw- on amd64 unless they created it differently. (The differences are part of history and part of the implementation [PLT tables are instructions on SPARC but are data on x86]) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>Which does the PT_SUNWSTACK header apply to? It sets the fields in the proc structure which defines the stack protection. >I'm not sure what point there would be in per-thread >control; if any one thread in an address space is exploitable, >the whole address space is potentially corruptible. It's different because mprotect() works on the process. Using mprotect requires you to run mprotect on all process stacks (as long as they are created by the library). I'm not sure that the kernel has sufficient information to figure that out. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>How does the trampoline work on architectures which have noexec bit >for stack by default, like sparcv9? A long time ago, I made some changes to gcc; I'm not sure if the changes are still in current gcc, but what happens is that we run mprotect() on the parts of the stack where we need to execute code. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>I picture this somehow as being just a bit more functionality added to >mprotect(2): > >/* following magic to identify operating on that segment, rather than > * a particular address > */ >#define ADDR_STACK (void *) (-1) >#define ADDR_HEAP (void *) (-2) > >mprotect(ADDR_STACK, 0, (PROT_READ|PROT_WRITE|PROT_EXEC)); It's a bit harder to bolt on the current implementation of the stack protection. And what is the "ADDR_STACK" *all* current thread stacks, the stack of main or the current stack? >where a length of 0 implies something similar to what (MCL_CURRENT|MCL_FUTURE) >implies with memcntl(2), namely to apply that behavior to both the present and >future pages of the segment; that would combine applying mprotect() to the >existing pages as well as setting p_stkprot or p_datprot in the proc structure. > >Such an interface would be ideal for runtime control either by >the developer or after-the-fact with an LD_PRELOAD'ed shared object. > >Adding this functionality to mprotect() would be more understandable than >a new function, and would avoid adding an additional system call. I don't >imagine that any existing software (except _maybe_ an emulator?) would >call mprotect() so often that the addition of a couple of if's or a switch() >applied to the addr arg would present a performance problem. There is one particular issue where gcc uses a trampoline created on the stack when you are passing a nested function as a function argument. (A nested function requires an additional argument) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>I had a few minutes today to try an experiment, and I'm afraid >the idea of having ld always generate a PT_SUNWSTACK is a non-starter. > >The problem is that it overrides the behavior of 'set noexec_user_stack=1' >in /etc/system, and can therefore quietly allow programs that would >not previously been able to execute on the stack do so. > Thanks for this investigation. There is another issue we haven't explored is the use a system call; there's a sysconf(_SC_STACK_PROT) but there's no way to set in on the fly. If we create a function to change it on the fly, we could make a LD_PRELOAD object which enforce it. The current mapped pages would not be protected but threadstacks and additional pages would be rw-. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] No network/Internet
>Here are the commands i've typed. And additionally, a image of my screen. > > >j...@opensolaris:~$ svcadm disable svc:/network/physical:nwam >j...@opensolaris:~$ svcadm enable svc:/network/physical:default > > > >j...@opensolaris:~$ ifconfig nge0 >ifconfig: status: SIOCGLIFFLAGS: nge0: no such interface Try "ifconfig -a plumb"; if the interface is supported, ifconfig -a will then show it. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>Thanks for the info. > >Doesn't help with existing binaries, but it's good to know for >when one can re-link. If there's a way to construct an >LD_PRELOAD'able shared object to do the trick for existing >binaries, I don't see how given the description of mprotect(2), >to specify the entire present (and future) stack or heap of a process. True. That's why we have the /etc/system tunable. Before we created the tunable, I created a "poke the kernel" script. Because of the limitations in where I could poke the kernel, I modified the routine which created all the "zero-fill-on-demand" pages; this includes the stack but everything allocated in on the heap, including the BSS segment. >From some of the stuff I've looked at, it seems that executables were >first widely built to disable an executable stack >(using something as boring as /usr/bin/who as the test case) >in Solaris 9; on Solaris 8, I didn't see a program header for the stack for >the "who" executable, but when I recompiled with the -M option you >gave, and checked the new binary with elfdump, I saw the additional >program header. I'm not sure when we added support for that header in the kernel. >There are a few oddball programs that legitimately want a segment >that's both writable and executable (because they compile on-the-fly); >some LISP interpreters used to do that, I think; and I can imagine that >some Forth implementations might want to do strange things. (I wonder >if the JVM does anything like that?) But I suppose they ought >to create an explicit anonymous mapping with the permissions they >desire for that use. With the patch script above, we did find that java and some other runtime create code run into the problem. But we limit it to the stack as the kernel does, there is no issue apart from some of the cases where gcc creates code on the stack. >Now...if there only weren't still 32-bit executables that aren't large-file >aware. :-) The only issue we were able to fix was the inability to open more than 256 FILE*s and we have a pre-loaded script. I don't think we have a routine to change the stack protection on the fly. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] 32-bit noexec_user_stack on per-process basis?
>ppgsz(1) or mpss.so.1(1) can set preferred page size for >existing (ppgsz) or new (both) processes. > >Is there anything that can similarly remove execute permission >from the stack of 32-bit processes, without do so on a system-wide >basis (i.e. without putting set noexec_user_stack=1 in /etc/system)? Without relinking, you mean? If you can relink the program, there's -M /usr/lib/ld/map.noexstk >It strikes me that something like that might be good with high risk >programs like browsers, that are often still 32-bit, and usually >have a shell wrapper anyway. We use this in many consolidations: % pmap `pgrep firefox`|grep stack E60FA000 8K rw--R[ stack tid=2846 ] F727A000 8K rw--R[ stack tid=14 ] F74F8000 16K rw--R[ stack tid=13 ] F7AF8000 16K rw--R[ stack tid=7 ] F83FA000 8K rw--R[ stack tid=2858 ] FA5FA000 8K rw--R[ stack tid=5 ] FA6FA000 8K rw--R[ stack tid=4 ] FABFA000 8K rw--R[ stack tid=3 ] FACFA000 8K rw--R[ stack tid=2 ] FFB9C000 400K rw---[ stack ] FFBFC000 16K rw---[ stack ] You can detect such programs using "elfdump"; you'll find a Program header similar to this: Program Header[5]: p_vaddr: 0 p_flags:[ PF_W PF_R ] <- RW, no N! p_paddr: 0 p_type: [ PT_SUNWSTACK ] p_filesz: 0 p_memsz:0 p_offset: 0 p_align:0 Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] // napp-it WebUI + NAS-appliance Online-Installer with OpenIndiana support
>Hi, > >yesterday I tried to set up the webservd as requested and I removed the >All:suser:cmd:::*:uid=0 from /etc/security/exec_attr, but after doing that I >couldn't access any low level system function in napp-it, like viewing the >discs or the zpools. All should be exacty like: All:suser:cmd:::*: If you change it, things will break. >usermod -P 'Primary Administrator' webservd didn't work either, >which worked in snv134. In oi147 i t barks like this to me: > >usermod -P 'Primary Administrator' webservd >UX: usermod: ERROR: Roles must be modified with ``rolemod''. So use rolemod. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris 11 Express
>Oracle does not own the complete Copyright on the code, so they need to follow >the CDDL. All the code where Sun added the CDDL are owned by Sun, now Oracle, *including* all the contributed files through SCAs. Which files are you referring to where Sun/Oracle doesn't have the copyright? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris 11 Express
>Uros Nedic wrote: > >> I'm wondering if it possible that Solaris Development could be op= >ened at least on a binary >> level so we could update on a bi-weekly basis our builds as we did = >with OpenSolaris? > >If Oracle did this, Oracle would be in conflict with the CDDL. No. >If they publish binaries, they need to publish sources as well. No, they don't need to publish the sources; they *own* the copyright on the source and they can do what they want. The CDDL license is a license for those who do not own the copyright. Nexenta, Schillix, etc, they are required to publish the files they use under CDDL. But Sun wasn't and neither is Oracle. (Even if the Solaris source was distributed under FSF, there was still no need for Sun/Oracle to publish the source when cutting a binary release). Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [illumos-Developer] zpool upgrade and zfs upgrade behavior on b145
>Additionally, even though zpool and zfs "get version" display the true and >updated versions, I'm not convinced that the problem is zdb, as the label >config is almost certainly set by the zpool and/or zfs commands. Somewhere, >something is not happening that is supposed to when initiating a zpool >upgrade, but since I know virtually nothing of the internals of zfs, I do The problem is likely in the boot block or in grub. The development version did not update the boot block; newer versions of beadm do fix boot blocks. For now, I'd recommend you upgrade the boot block on all halves of a bootable mirror before you upgrade the zpool version or the zfs version. export/import won't help. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] zfs send/receive?
>hi all > >I'm using a custom snaopshot scheme which snapshots every hour, day, >week and month, rotating 24h, 7d, 4w and so on. What would be the best >way to zfs send/receive these things? I'm a little confused about how >this works for delta udpates... > >Vennlige hilsener / Best regards The initial backup should look like this: zfs snapshot -r exp...@backup-2010-07-12 zfs send -R exp...@backup-2010-07-12 | zfs receive -F -u -d portable/export (portable is a "portable" pool; the export filesystem needs to exist; I use one zpool to receive different zpools, each in their own directory) A incremental backup: zfs snapshot -r exp...@backup-2010-07-13 zfs send -R -I exp...@backup-2010-07-12 exp...@backup-2010-07-13 | zfs receive -v -u -d -F portable/export You need to make sure you keep the last backup snapshot; when receiving the incremental backup, destroyed filesystems and snapshots are also destroyed in the backup. Typically, I remove some of the snapshot *after* the backup; they are only destroyed during the next backup. I did notice that send/receive gets confused when older snapshots are destroyed by time-slider during the backup. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Merge other distros to OpenIndiana?
>> Cloning somewhat helps with the disk footprint, but >> the savings there will >> get lower over time as more and more binaries are >> upgraded. We clearly >> have to work on that. The memory saving of sparse >> root zones was nothing >> to sniff at. > >So a share-able executable, library, or object had >its read-only parts only resident once across the global and >multiple sparse zones (sharing a copy of the same object), >but now (even if the zone resides in a ZFS clone) is loaded >once for each (non-sparse) zone? Correct. Some may say that "dedup" will fix this but the memory footprint of dedup itself is huge. The first implementation of dedup didn't allow sharing memory pages even when the underlying storage was the same. >I bet that will be a big problem on for example a workstation >using Trusted Solaris extensions, where each label corresponds >to a zone; since the set of libraries for GUI applications tends >to be rather large (esp. with GTK+/GNOME as compared to the old >days of Motif). >Probably affects CPU cache, too. And the page tables. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Merge other distros to OpenIndiana?
>>> * SVR4 packages vs IPS vs whatever packaging scheme Nexenta uses >>> (there too there's a problem, inasmuch as other tools like beadm and zonecfg >>> are also involved, although I gather it ought to be possible to come up with >>> different versions of them for each packaging scheme that some distro or >>> another uses) >>> >> SVR4 packages support sparse zones, IPS does not. >> >> >OpenSolaris doesn't have sparse zones, so there nothing to support. > >If I remember the threads at the time, they were (rightly) considered >unnecessary now we use ZFS for zone roots. > I'm not sure I agree with "rightly"; a sparse root zone had several advantages which have been lost such as disk footprint but more importantly, shared executables, libraries, etc. Cloning somewhat helps with the disk footprint, but the savings there will get lower over time as more and more binaries are upgraded. We clearly have to work on that. The memory saving of sparse root zones was nothing to sniff at. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NetApp and Oracle settle patent dispute over ZFS
>casper@sun.com wrote: > >> Also, the law should be changed to disallow patent trolls: if you don't >> ship or plan to ship a product using your patent, then clearly you're not >> suffering when someone else is using a similar invention. > >This is against the original intention for patents: > >Patents have been filed by inventors that do not product products and that look >for a manufacturer... Sure, but practice has changed; now patents are sold to patents trolls and not to industry. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] NetApp and Oracle settle patent dispute over ZFS
>There's also a public good argument. When you patent something, yes, you get >a temporary monopoly -- but the tradeoff is you have to publicly reveal it. >In the absence of patents, more technological advances would be treated as >trade secrets and kept out of the public eye. > >I do think there are a lot of abuses of the patent system, most of them >related to insufficient vetting of new patents and the expense of litigating >them later. I think something like the patent system is necessary, though. > The length of the monopoly should be part of the discussion; I can see why a patent for a medication should be as long as it is now, but for software and perhaps hardware too, I'm thinking more of 5 years and not longer. Also, the law should be changed to disallow patent trolls: if you don't ship or plan to ship a product using your patent, then clearly you're not suffering when someone else is using a similar invention. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] kudos
>You'll notice something similar, once you point your MS Windows based brows= >er >to a CIFS server that's build using ZFS and ZFS-snapshots... > >MS is also capable of interpreting the CIFS versioning/ZFS snapshot/history= >... Thanks for pointing this out; I'm using my Solaris server as a CIFS home server and I never noticed the "View previous versions". Wonderful. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [illumos-Developer] SchilliX-0.7.1 based on b147+ available
>On Tue, Aug 31, 2010 at 01:10:23PM +0200, Joerg Schilling wrote: >> SchilliX-0.7.1 is based on the latest ONNV source published by Oracle. >> This is build 146 with 16 additional putbacks. >> >> see: ftp://ftp.berlios.de/pub/schillix > >I just tried booting this image on my Acer Aspire AspireAS3810T-6827 >laptop, in hopes that the IOMMU code in the kernel had been improved >since OS 134. Unfortunately, it seems to have the same problem. The >last few messages on the screen were: > >ramdisk0 at root >ramdisk0 is /ramdisk >root on /ramdisk >acpinex0 at root >acpinex0 is /fw >acpinex: c...@1, cpudrv0 >/fw/c...@1 (cpudrv0) online >pseudo-device: dld0 >dld0 is /pseudo/d...@0 > >At that point it hung. I had to hold down the power button to >recover. Windows 7 and Ubuntu run fine on this laptop. Is there >anything else I can do to diagnose the problem? Boot in kmdb and try to find out where it hangs. It didn't hang before? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris cancelled, to be replaced with Solaris 11 Express
>On 08/19/10 10:09 AM, usafverteran wrote: >> What's your point? IBM has had DB2 Express for quite some time. >> > >The point was the statement "Name one downloadable Oracle product which >is limited in some way. (Other than the license)". > >Dmitry simply found one example. He was right; it's clearly hard coded into the database. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris cancelled, to be replaced with Solaris 11 Express
>"Dmitry G. Kozhinov" wrote: > >> >I've never heard of any version of Solaris, commercial or otherwis= >e, >> >that was purposely crippled to use only limited RAM, disk space, >> >and CPU cores. >> >> That's because Solaris was under Sun, not under Oracle. > >In 1990 Sun did tell costomers that they have been planning to bundle the OS >license to the hostid of a machine. Fortunately, Sun did listen to their >customers and did not make it reality. > And why did the OP expect that Oracle will behave differently: "Please note that Oracle doesn't use software keys. You can just install the software and use it. It is up to you and your consciences to license the software before using it. " (from the orafaq website; of course, it's not from the horse's mount) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris cancelled, to be replaced with Solaris 11 Express
>> Let me guess what the Solaris 11 Express will be: >> >> The demo version of commercial Solaris 11. >> Utilizing only limited amount of RAM, disk space, and >> limited number of CPU cores. > >I've never heard of any version of Solaris, commercial or otherwise, >that was purposely crippled to use only limited RAM, disk space, >and CPU cores. Nexenta? Or is that using the honor system, again? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris cancelled, to be replaced with Solaris 11 Express
>Let me guess what the Solaris 11 Express will be: > >The demo version of commercial Solaris 11. >Utilizing only limited amount of RAM, disk space, and limited number of CPU >cores. Name one downloadable Oracle product which is limited in some way. (Other than the license) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] libmtmalloc vs libumem
>Hi, >I am able to understand how libmtmalloc works from the documentation of >libmtmalloc.c source file. However, I am unable to find proper documentation for libumem. Could someone provide the key diffe rences between libmtmalloc and liumem, please? Please also provide me links to the documentation/pa per based on the design of libumem. > Libumem is the userland implementation of Jeff Bonwick's slab allocator; you should be able to find the paper he wrote using google: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.29.4759&rep=rep1&type=pdf Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] So who is ready to be let down?
>And the other operating systems I printed release dates for are RELEASES. >I'm not talking about development releases like you mentioned. If I were >to do that the list would be too long. All of those are RELEASES. OpenSolaris >hasn't had a RELEASE in 13 months (and it's 14 months old). z/OS has even had >2 releases since the last OpenSolaris RELEASE. AIX has v7 soon, but has had >a number of updates for 5.3 and 6.1. So why are you discounting Solaris releases other than OpenSolaris? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] samba version
>So, basically with out clear communication from Sun, Sun has abandoned a >key functional component? > Yes, better ACL integration would be nice but to use CIFS I have to setup >a WIndows domain controller? I use CIFS at home and I don't have a AD controller. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Netservices Limited
>Does anybody know of a way to exclude a service that netservices limited would >stop. > >I'm thinking along the lines of having the ftp service running all the time >even >if netservices limited is executed. I know I can start the service afterwards >but I would like to if possible exclude it. netservices limited is designed to use it *once* after a system was installed; the default install is "netservices limited" so you actually never need to run "netservices limited"o on OpenSolaris (but you need to run it on Solaris 10) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] pam_list and pam_sm_authenticate errors - console login opensolaris 2009.06
>Dear All, > >Having enabled pam_list in /etc/pam.conf I can then happily control who can >log on via ssh and that works fine. > >HOWEVER, once pam_list is included in /etc/pam.conf then console logins ALL >fail with messages such as > > >Apr 30 11:32:55 phoenix login: [ID 825731 auth.error] dlsym failed >pam_sm_authenticate: error ld.so.1: login: fatal: pam_sm_authenticate: can't >find symbol > > You didn't properly use pam_list. pam_list(5) says: other account requisite pam_roles.so.1 other account requiredpam_unix_account.so.1 other account requiredpam_list.so.1 specifically it is important to use "account" and *not* auth. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Oracle Corporation SunOS 5.11 snv_138 Apr. 15, 2010
> >Oh look ... snv_138 appears quite suddenly out of nowhere .. cool. > >Thanks to whomever made that happen. > >Oh, I see one of the big sources of delay. Its good to see the branding >process is going well. > >r...@aequitas:/root# uname -a >SunOS aequitas 5.11 snv_138 i86pc i386 i86pc Solaris > >r...@aequitas:/root# cat /etc/motd >Oracle Corporation SunOS 5.11 snv_138 Apr. 15, 2010 >SunOS Internal Development: gk 2010-Apr-15 [onnv_138-tonic] >bfu'ed from /build/archives-nightly-osol/i386 on 2010-04-30 >Sun Microsystems Inc. SunOS 5.11 snv_130 November 2008 > >The name "Sun Microsystems Inc." lingers on in the motd. > You've bfu'ed so part of the new motd is the motd of the older release. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] URGENT: Kernel lets 32bit process consume >4G memory?
>A 32-bit process _can't_* be bigger than 4GB, and as far as the kernel >is concerned, AFAIK won't be bigger than 2GB in terms of regular memory >(although it could have a frame buffer or something mapped in the >part of the address space reserved for I/O devices, making its total >size perhaps appear larger than 2GB, but still definitely <= 4GB). I can be nearly 4GB but not over 4GB. (On a 64 bit kernel or on sparcv9 there is no "kernel address space" as part of the userland address space) >*actually, a 32-bit _address space_ can't be. On suitable hardware, a 32-bit >kernel >can use special instructions to address more than 4GB of RAM, and some other >OSs allow even a user process to own more than one address space. I don't think >any of that applies here though. We had a specific form of memory based filesystems for x86 and it is also possible for a 32 bit process can cache more then 4GB in the file cache. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Review of OpenSolaris at Desktop Linux Reviews
>casper@sun.com wrote: > >> >The SVR4 kernel was developed from SunOS-4.0, but SunOS-5.x (Solaris-2.x) >> >was developed from SunOS-4.1.x. >> > >> >> I don't think that that is correct; SVR4 was developed with AT&T from >> SunOS 4.x and SVr3; the SCCS files proof that. Unfortunately, the code >> was copied as bare source and then check in at Sun. Solaris 2.x was >> clearly derived from SVr4 but a lot of changes were made. > >Do you have SCCS files from the time between 1987 and 1990? Yes. >I have a SVR3 AT&T source from 1986 and a S5R4V1 source from January 1992. >This allos to verify, that the SVR4 kernel is _very_ close to the SunOS-4.0 >kernel and very far from the S5r3 source. Possible; but much of the Solaris userland was derived from SysV and not from SunOS 4.x. >Even the S5R4V1 source did not include any of the new features in e.g. the UFS >implementation, such as "clean flag" and I/O clustering. > >Also S5R4 neither comes with "lofs" nor with "hsfs". > >Later S5R4 versions added "cdfs" as a "hsfs" replacement. Certainly possible; it is not nlikely that both AT&T and continued developing in parallel while they were merging parts into SVr4. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Review of OpenSolaris at Desktop Linux Reviews
>bsd wrote: > >> The way he states it in his article is that Sun made significant changes = >after buying SVR4, which is incorrect, because changes to SVR4 were made af= >ter Solaris was open sourced. Significant changes were made when SVR4 was = >developed, not after. Solaris 2.0 was developed from SVR4. > >The SVR4 kernel was developed from SunOS-4.0, but SunOS-5.x (Solaris-2.x) >was developed from SunOS-4.1.x. > I don't think that that is correct; SVR4 was developed with AT&T from SunOS 4.x and SVr3; the SCCS files proof that. Unfortunately, the code was copied as bare source and then check in at Sun. Solaris 2.x was clearly derived from SVr4 but a lot of changes were made. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Seeing .nfs0000000000002def00000002 on nfs mounts from linux
>Using an Osol b 133 nfs server. When working from a linux client I >sometimes see files like: .nfs2def0002 > .nfs2f6b0001 > >Appear on the filesystem as I'm working... any attempt to remove them >(from the linux client) fails, with the message: > > (file-error Removing old name Device or resource busy > /projects/reader/perl/mapping/.nfs2f6b0001) > >What are these? When a file is unlinked by an NFS client but the file is still open, the client actually renames the file and later removes it after the file is closed. (It is created by the client and it shoul be removed by the client; the client makes sure you don't remove it) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] toshiba laptops opensolaris preinstalled
>On Tue, 20 Apr 2010 07:45:01 PDT, Edward Martinez >wrote: >> I was wondering if i would be voiding the laptop warranty by erasing > >> windows and installing opensolaris either on a m10 or a portege ? >> >> > >Warranties for general-purpose computer hardware that dictate what >software you can install are unheard of! It would be a sad day for our >rights if they come to pass. It happened before (certain hardware will fail if the OS doesn't run the fans, step down the clock, etc) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Is rolemod only for uid < 1024 ?
>Hi > >I get an error > ERROR: Roles must be modified with ''rolemod''. >when running /usr/sbin/usermod -c 'Super-User' -d /sub_home/root -e ' >' -f 0 -g 0 -u 0 -s /usr/bin/ksh root > >And Users must be modified with ``usermod''. > When running /usr/sbin/rolemod -c 'Default tut' -d /export/home/tut >-e ' ' - f 0 -g 1000 -u 1091 -s /bin/rksh tut > >On an snv_134 X86. > >Is rolemod for uid < 1024 >and usermod for uid >=1034 > >I'm not able to deduce this from the man entries on both manuals. "rolemod" can be used for "roles"; "roles" are users with "type" set to role in /etc/user_attr. "usermod" can be used for ordinary users; they don't have a type attribute or a type set to "normal". Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] brtfs on Solaris? (Re: [indiana-discuss] So when are we gonna fork this sucker?)
>brtfs could be supported on Opensolaris, too. IMO it could even >complement ZFS and spawn some concurrent development between both. ZFS >is too high end and works very poorly with less than 2GB while brtfs >reportedly works well with 128MB on ARM. Both have license issues; Oracle can now re-license either, I believe, unless brtfs has escaped. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [indiana-discuss] So when are we gonna fork this sucker?
>Just a completely different question...is there any plans for btrfs ? >Will ZFS and btrfs co-exist or there's a chance that the less used one >would be dropped? Which OS supports both? Linux support brtfs, Solaris supports ZFS. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] can't edit files with vi properly
>Not sure, if this is the correct group. >I'll give it a shot anyway. > >sshing into the gloabl zone and editing a file there with vi (TERM set to >xterm) is no problem. >However- zlogin to a zone (TERM is xterm here too - set by .profile) WYSINWYG. >Only when I position the curser to the left column and move right, the real >contens of the line ov errides the previously (wrong) displayed. >You may imagine that this is a tedious editing and I'm sure, I'm not the first >person haveing the problem. I tried setting term to vt100 too - didn't help. Set term to xterm. Run "eval `resize`" after logging in. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Any one knows which package libzonecfg.h is in and how to find out?
>Hello > >Does anyone know which package libzonecfg.h is in and how to find out? > >I'm not able to find the file in sxce 127. > >Any ideas are welcome Part of the source, not distributed. Why do you need it? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] syslog-ng?
>casper@sun.com wrote: >> Clearly syslog-ng needs to be able to pick up such messages *and* create a >> door where it can say "yes, I'm still running". > >Wouldn't it be better to use SMF for that now instead of a custom door >solution for just that one service? Considering that we need to call that door *every time* we log a message, wouldn't that be *really* expensive? I'm afraid that will pull in a few extra libraries into libc. The current code is broken: /* output the message to the local logger */ if ((putmsg(LogFile, &ctl, &dat, 0) >= 0) && syslogd_ok()) return; if (!(LogStat & LOG_CONS)) return; The call to syslogd_ok() should be done *after* checking that LOG_CONS is set and the putmsg() failed. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] syslog-ng?
>Saw someone in opensolaris-help having trouble building syslog-ng, >and it got me to wondering why that hasn't been integrated (perhaps >with any changes needed to support doors rather than a Unix domain >socket for local communication). > >AFAIK, the existing syslog has a lot less functionality. Aside from >using doors for local communication, does it have any _more_ functionality >than syslog-ng? syslog does not use doors to log messages; it uses the door to find out whether syslog is still up and running. It logs the messages in some stream device. (/dev/sysmsg or some such). Clearly syslog-ng needs to be able to pick up such messages *and* create a door where it can say "yes, I'm still running". Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] AMD64/NUMA: is it possible to limit amount of memory available to one CPU?
>Richard L. Hamilton wrote: >>> C, the standard network stack runs on CPU0 (along >>> with many other >>> kernel items), though it might need some tuning for >>> your particular app. >>> >> [...] >> >> I thought that at least since FireEngine, the network stack could >> use multiple CPUs! >> >Oooh! I just looked at the FireEngine docs, and while I was aware that >we'd broken the all-NICs-need-CPU0, I hadn't bothered to read the >FireEngine implementation. The Solaris network stack never used one CPU or specifically CPU0. Not now, not ever. (Well, not since SunOS 4.1.x) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris 10 - no longer free
>Mike, > >> I do not notice anyone from Sun/Oracle making any comment on this thread. >> >> Why are you guys so quiet? >> > >Aside to AlanC's (always) measured response, I suspect an implicit >gagging order might be present. This is just too funny. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] MythTV
>On 03/18/10 10:20 AM, homerun wrote: >> Greetings >> >> Just asking , is there interest to port MythTV to OpenSolaris. >> http://www.mythtv.org/ >> >How long will the supported list of video capture cards be useful? >The notices I've been receiving from Comcast sound like they are >moving to most programming being delivered as encrypted digital. >If this an indication of where the industry is moving, how useful >will the boards be that only support over the air (ATSC) or >unencrypted cable (clear QAM)? Same here; there are several Cable companies in the Netherlands; one stops analog service on 10/10/10; the other two are larger but I expect them to stop with analog service in 5 years or so. You will need a digital TV receiver with a CI+ slot (and a cable operator who supports the CI+ slots) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Cannot boot after power failure...
>On Sat, Mar 20, 2010 at 12:34:15PM +0100, casper@sun.com wrote: >> >> >Thanks for that explanation of why there is no failsafe boot option for >> >OpenSolaris. I had wond ere >> d about that, since I'd seen it in Solaris Nevada/Express. >> > >> >> Ok, except that the CD boot is much slower. >> > >> >Why is that a problem? Why should one worry about the boot time for a >> >rescue procedure? Most users of OpenSolaris are unlikely to ever need to >> >employ such a procedure. >> >> Really? The boot-archive isn't always updated when it is needed and I had >> this happen often until I "nulled" the boot-archive script. > >Can you explain a bit more about what you mean by "nulled" the >boot-archive script? In some systems I removed the script and changed it to "exit 0" but mostly I change the startup script to :true in the SMF configuration. I've been bitten several times, years ago, and since then I always "null" the boot-archive start script. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Cannot boot after power failure...
>Thanks for that explanation of why there is no failsafe boot option for >OpenSolaris. I had wondere d about that, since I'd seen it in Solaris Nevada/Express. > >> Ok, except that the CD boot is much slower. > >Why is that a problem? Why should one worry about the boot time for a >rescue procedure? Most users of OpenSolaris are unlikely to ever need to >employ such a procedure. Really? The boot-archive isn't always updated when it is needed and I had this happen often until I "nulled" the boot-archive script. This problem should not have happened it is completly avoidable. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Cannot boot after power failure...
>I have a storage server running on snv_129. When trying to reboot my server >it was stuck in the perpetual loop of the happy face boot screen. After> trying to boot in single-user mode, I got this error message: > >WARNING: The following lines in / differ from the boot archive: > changed /etc/devices/devid_cache > changed /etc/name_to_major > changed /etc/path_to_inst Can you edit the boot argument by adding a "-s"? Then you should do the following: svccfg -s boot-archive svc:/system/boot-archive> setprop start/exec = :true This fix the problem (typically, the boot archive is usable; it needs to be updated when: - the time zone changes - when you add/remove a USB/Firewire device - install a device driver Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris 2010.03 improvements?
>Assuming you mean Firefox, remember that (Open)Solaris x86 runs on >both 32-bit and 64-bit with a single binary distro (unlike Linux, where AFAIK >a 64-bit kernel normally runs only 64-bit binaries, so 32-bit and 64-bit are >separate binary distros*). So to have 64-bit Firefox and still be able to run >Firefox on 32-bit, there would need to be two copies of everything: Firefox, >bundled plugins, etc. Not to mention every single one of the many shared >libraries that it depends on. Many of the plugins are not available and, e.g., only an alpha version of 64 bit flashplayer is available for Linux. >Yes, on x86, a 64-bit binary is usually a little faster (not to mention being >able to address _much_ more memory). But is it enough faster to be worth >having two copies of something as big a browser? Or do you have so much >RAM on your box that having the browser grow past 2GB address space is >no big deal? I'd prefer to my firefox to die before it has leaked 4GB :-) So why do you really want a 64 bit firefox binary? "Mine goes to 11". Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] about zfs exported on nfs
>I've had bad experience setting NFS mounts in /etc/fstab. The problem is: >If the filesystem can't mount for any reason, then the machine doesn't come >up. Unless you set it as a "soft" mount, in which case, the slightest >little network glitch causes clients to lose their minds. There is also a "bg" mount option: the mount will continue in the background when it fails; however, if a NFS mount is always needed, I suggest to create it in /etc/auto_direct. The mount isn't performed at boot then but it is delayed until the mountpoint is accessed. >What I wrote in the previous email, about using automount and hard >interruptable NFS mounts was very well thought out and based on years of >commercial deployment of NFS systems. Like I said, it's rock solid if >configured as I described. It's resilient against network failure during >boot, or during operation, yet it's force-interruptable by root if >necessary, which is extremely rare. I agree. automount and not /etc/vfstab. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Upgrade 133 -> 134, now can't boot anything :-(
>Thanks for the suggestions, but I think this is my own fault. I think >I've shot myself in the foot by enabling compression on rpool. > >So: `root (TAB' >gives you a nice clue about how to proceed, it says there's a ZFS >filesystem at partition 2. >`findroot (pool_rpool, 2, a)' >succeeds, and finds me a root filesystem >`cat /boot/grub/menu.list' >finally tells me the problem: "unsupported compression algorithm" (or >along those lines, sorry, this is from memory, I can't capture the output). >So I think I need to boot from a DVD of a sufficiently current release >to understand the zpool/zfs versions, import the pool, turn off >compression, trash the (compressed) new boot environment, activate the >previous environment, and hope I'm back on track. I do think the system >should have told me that I was making my rpool unbootable by turning on >compression, though, and made me use a `-f' flag or something ZFS should not allow you to enable the wrong compression. You don't need to trash the rpool, but you mount it from the liveCD (or some other boot environment). Change the compression algorithm (the default will work). Tar and unpack the imported files. As soon as the kernel is booted, anything compressed will work; I think it might be sufficient to rebuild the boot-archive. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [desktop] NVIDIA driver install
>Thank you! Hum... >j...@opensolaris:~$ pfexec beadm list >BEActive Mountpoint Space Policy Created >---- -- - -- --- >opensolaris - - 20.48M static 2010-03-05 14:00 >opensolaris-1 - - 13.75M static 2010-03-05 16:11 >opensolaris-2 - - 313.49M static 2010-03-05 18:37 >opensolaris-3 NR / 6.36G static 2010-03-09 13:22 >j...@opensolaris:~$ pfexec beadm create opensolaris-3-nvidia >j...@opensolaris:~$ pfexec beadm mount opensolaris-3-nvidia /mnt >j...@opensolaris:~$ pfexec pkg -R /mnt uninstall NVDAgraphics >Creating Plan >pkg: 'NVDAgraphics' matches no installed packages > >j...@opensolaris:~$ pfexec pkg -R /mnt uninstall NVDAgraphicsr >Creating Plan >pkg: 'NVDAgraphicsr' matches no installed packages > >j...@opensolaris:~$ pfexec /bin/sh NVIDIA-Solaris-x86-173.14.25.run >--extract-only >j...@opensolaris:~$ cd NVIDIA-Solaris-x86-173.14.25 >j...@opensolaris:~/NVIDIA-Solaris-x86-173.14.25$ pfexec pkgadd -R /mnt -d >NVDAgraphics NVDAgraphic sr >pkgadd: ERROR: no packages were found in > > >Where is my fault...? You need to specify the directory (current directory) pfexec pkgadd -R /mnt -d . NVDAgraphics NVDAgraphic ^ - missing argument Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Performance comparison of OpenSolaris upgrade vs liveupgrade
I've been upgrading a new systems and compare them with SXDE (118 or later) Output: "time, I/O done" Ultra 40, sufficient memory SXDE 28m16 read 924M write 6192M OSOL 29m44 sd1 read 4305M write 2243M M3000, sufficient memory SXDE (+/- 16 minutes, between the upgrade_script was created and the upgrade was done) OSOL 22m23 sd0 read 1218M write 2895M We know that pkg uses a lot of memory and so this is what happens when you run it on a system with only 1GB: (yes, that is: 1 day, 18 hours), Ultra 60 OSOL 1d 18h13.12 sd0 read 18G write 6624M I/O to swap: sd2 read 187G write 191G (!) (depending on the amount of change, the Ultra 60 ran liveupgrade around 1h40m) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] very long wait in the end of pkg install
>It only *looks* like it downloads everything again -- this is purely a >misleading display issue that we're hoping to improve at a future point. It did take a reasonable amount of time. Not less than downloading the first time, but I can't be sure. The ips server is local, though. >What happens is that the transport system iterates through all the >content needed for the operation, verifies that it's valid, and then >skips the download for the file because it already has the content. > >If you actually snoop network traffic, you'll see it doesn't actually >download all the files again. >>> I think you're confused about what I've said here. And no, the data is >>> intentionally stored by digest and not by its original name. It would >>> neither be practical, nor efficient to store the files in the repository >>> by their delivered name. The size of the /dev repository is currently >>> around 63.6G and growing... >> >> A hardlink doesn't cost (much) of additional storage. > >Since files are retrieved by digest, there's no reason to cache them by >any other means. But it would be useful. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] very long wait in the end of pkg install
> > >>For example, on my system, when I did an image-update from build 110 to >>build 111, with all of the data to be installed downloaded already, it >>only took about five minutes to upgrade 666 packages and move around 480 >>megabytes worth of data. That's not too shabby if you ask me. > >Compare upgrading from build 128 to later :-) Sorry, I meant from build 118; the continuous rewriting of the contents file was a bug which was easily fixed. I don't think you should benchmark against a completely broken system; to use that then your bar would by very low. What was interesting is that "zfs" improved the old contents file issue on systems with more than 1GB of memory; with "ufs" each update of the contents file was completely written to disk; this one done at least twice for each package. With zfs much of the contents file was never written to disk. On laptops, with slow disks, the difference would be greater. During the development I noticed how much I/O was done before/after when installing a zone: 15m06 sd1 read 59M write 34G (ufs) 6m30 sd1 read 66M write 583M (zfs eats 33GB of I/O) but after it sis more like: 4m02 sd1 read 83M write 319M (ufs) 2m59 sd1 read 76M write 129M A full install used to write 150GB. The use of "pkzip7" further penalizes the old install; uncompressing takes around 1/3rd of the install time. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] very long wait in the end of pkg install
>For example, on my system, when I did an image-update from build 110 to >build 111, with all of the data to be installed downloaded already, it >only took about five minutes to upgrade 666 packages and move around 480 >megabytes worth of data. That's not too shabby if you ask me. Compare upgrading from build 128 to later :-) >Except that pkg(5) actually does both. It keeps the package data >spooled locally, and if the user has chosen to purge the download cache, >it can retrieve just the data it needs. So when "upgrade image" dies, e.g., because there's a symlink or directory in the "wrong place" and you remove the offending file, why does it download everything again? >I think you're confused about what I've said here. And no, the data is >intentionally stored by digest and not by its original name. It would >neither be practical, nor efficient to store the files in the repository >by their delivered name. The size of the /dev repository is currently >around 63.6G and growing... A hardlink doesn't cost (much) of additional storage. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] very long wait in the end of pkg install
>> * efficient update operations >> >> -- Because pkg(5) only retrieves the files that have changed between >> package versions between updates, the client only downloads exactly >> what it needs to perform update operations. Other systems have chosen >> to implement this by pre-generating deltas between package versions, >> but that also means that their users have to rely on a pre-generated >> delta being available for every possible origin point from which they >> need to update from and to. Do we actually implement that? Currently, every new release revs all the binaries; how much of the OS is not a binary? (Upgrading using "pkg" is much slower than "liveupgrade"; in both cases I use a local server). >> * efficient repair operations >> >> -- Because pkg(5) only has to retrieve the individual files it needs >> to perform an operation, this greatly reduces the time needed to >> restore missing package content on a system. The latency will nearly always kill you; I believe that the Windows method is superior: you do keep all the package spooled locally but repairing is than simple and can happen off-line. >> * greatly reduced publication resource costs >> >> -- Because pkg(5) stores package content as individual files (by >> content hash), files that are identical between packages are shared on >> the server. In addition, files that don't change between package >> versions are also shared. This can greatly reduce the resource-cost >> of publication storage and publication time. While I'm fine with that, I would suggest that you also use hardlinks and also keep the data installed linked to the proper name. E.g., I have broken "python" and now I need to copy the data from the server. Where is it? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] very long wait in the end of pkg install
>On Sat, Feb 27, 2010 at 12:48 PM, wrote:
>>
>>>well the whole install took 2 hours
>>>for 1 pkg !!
>>>it seems that disk read/write here is the bottleneck (?).
>>>iostat -D 2 shows my hd busy at around 100%
>>>iosnoop reports a lot of access by pkg :
>>
>> How much memory? =A0It is possible that you're paging to death.
>
>
>
>Hi Casper,
>
>I am subscribed to pkg and caiman.
>
>Maybe his system is swapping to death.
>But why??? If you compare memory- and CPU- usage of IPS to those
>values of *every* other system on the market, it is unbelievable at
>the first look. ^Horrific^ - to find the right term.
Yes, that is true.
I'm not sure why it needs so much memory; it is possible that using
python contributes to the memory use. Both perl and python of wonderful
primitives but you pay for how the primitives are implemented: they don't
give you a choice.
A long time ago, we re-wrote the "contents" file database in a (SQL)
database; that was a bad idea for several reasons: you have again no choice
on how the data is stored and this again added to huge memory pressure
and apart from other issues such as not properly sorting the contents file,
it made installing much slower and not faster. The contents file was then
around 15-20MB, the database used 512MB or more.
It took me around 10 days to re-write the svr4 package tools using some
form of home-grown database: but this database only needed indexing on one key,
the pathname, to me it was important to keep the memory use down as I
wanted to run it on small systems (upgrade with a shrunk miniroot on a
256MB system)
I used avl trees and stored the following:
typedef struct pkgentry {
char *line; /* The contents line for the file */
avl_node_t avl; /* The avl header */
int pkgoff; /* Where the packages live; start with SP */
int pathlen;/* The length of the pathname */
int len;/* Length of the line (incl NUL) */
} pkgentry_t;
I keep the whole line and I don't actually parse it and most of the work is
still done by the client programs.
The contents file was at the end of SXDE around 40MB(!); the daemon needs
only 60MB.
I don't know how much data the IPS needs to keep in memory, but I'm sure
that the penalty is much larger than the 50%.
It gets worse when you want to install/upgrade zones: (memory + cpu)* #zones.
You can scale if your operation is O(1) or the constant is small.
Casper
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] very long wait in the end of pkg install
>well the whole install took 2 hours >for 1 pkg !! >it seems that disk read/write here is the bottleneck (?). >iostat -D 2 shows my hd busy at around 100% >iosnoop reports a lot of access by pkg : How much memory? It is possible that you're paging to death. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Missing drivers in OpenSolaris.
I've noticed one particular is absent from OpenSolaris; this was a bit unfortunate for the SF 6800 I was upgrading. What other drivers are currently not distributed with OpenSolaris? I'm sure there's a list but I couldn't find it. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] SUNWluu packages for b132
>They are not available in IPS. You can install the SVR4 packages on your >system, but I am not sure what you need the live upgrade user package >for. We don't use the old live upgrade on OpenSolaris, we have 'Snap >Upgrade', which is managed via beadm(1M). I don't think they'll work properly either. If you have both an IPS and a SXDE environment you can change fairly easily. "beadm" will list zfs liveupgrade environments: BE Active Mountpoint Space Policy Created -- -- -- - -- --- osol_132 NR / 7.58G static 2010-02-15 12:04 snv_130 - - 5.18G static 2010-02-15 09:20 "beadm" misses some things needed for easily scripting (such as the "lucurr" command) but it can allows booting from an older environment. You can't boot a "liveupgrade" environment, you'll need to change the grubmenu (x86) and the bootfs in your zpool. I'm currently upgrading all my "liveupgrade systems" to "Opensolaris"; I can do this more or less automatically, except for editing the grub menu. There were some issues, such as the "isp" device driver not available in OpenSolaris (I need it to boot my SF 6800) It took me a few hours over the weekend to create and test the scripts. My mechanism to upgrade from SXDE to OpenSolaris is: - make sure you have rpool and you can boot from that environment; it needs to have room for two entire boot environments. - snapshot & clone "/" - save all the "volative/editable" files - remove all the sxde packages using "pkgrm" (removing all the post/pre remove scripts and all the r. scripts as they don't always work properly - rm etc/svc/deathrow - install the SVr4 packages for IPS (download ips gate, compile it) - create an image (pkg image-create ...) and install all the packages - copy all the system files from the current root to the IPS root (including your repository) - change bootfs in rpool - reboot - and you'll see some issues about CDE missing but these are then gone from the repository. Issues I ran across are: - IPS will die when something "weird" happens (e.g., if you still have a /usr/openwin *directory* then IPS bails out and needs to be restart after removing the offending file; then it will need to download everything (???) - IPS just install "everything"; you most for install SUNWcsd else other packages will fail because /etc/group doesn't exist. - IPS uses quite a bit of memory, systems with less then 2GB will be uncomfortable when running IPS. - when I made an error and I wanted to the packages then all directories with some local content were moved to var/adm/pkg/ lost+found (this is where /usr/local was moved to) - error reporting is not always helpful - building the recent ips gate seems to install some files in the wrong location. If you run SXDE from ufs, then, unfortunately, you will need to convert. Liveupgrade allows that (as I have done with most of my machines). If your root is a mirror, then it's probably even easier: just split the mirror and use the second half. I think you'll need at least 20GB for the root (two ZFS boot environments, with compression, most everything installed) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] strange ps behaviour
>On second thought it seems the ps command is switchning between bsd behaviour
>(old /usr/ucb/ps) if
options are not preceded with '-' and "new" solaris behaviour if there is a '-'
>Is that correct ?
Almost.
This is what the main routine in ps look like today:
int
main(int argc, char **argv)
{
const char *me;
/*
* The original two ps'es are linked in a single binary;
* their main()s are renamed to stdmain for /usr/bin/ps and
* ucbmain for /usr/ucb/ps.
* We try to figure out which instance of ps the user wants to run.
* Traditionally, the UCB variant doesn't require the flag argument
* start with a "-". If the first argument doesn't start with a
* "-", we call "ucbmain".
* If there's a first argument and it starts with a "-", we check
* whether any of the options isn't acceptable to "ucbmain"; in that
* case we run "stdmain".
* If we can't tell from the options which main to call, we check
* the binary we are running. We default to "stdmain" but
* any mention in the executable name of "ucb" causes us to call
* ucbmain.
*/
if (argv[1] != NULL) {
if (argv[1][0] != '-')
return (ucbmain(argc, argv));
else if (argv[1][strspn(argv[1], UCB_OPTS)] != '\0')
return (stdmain(argc, argv));
}
me = getexecname();
if (me != NULL && strstr(me, "ucb") != NULL)
return (ucbmain(argc, argv));
else
return (stdmain(argc, argv));
}
E.g.:
% ps -uxg
ps: unknown user xg
% /usr/ucb/ps -uxg
USER PID %CPU %MEM SZ RSS TT SSTART TIME COMMAND
...
% ps -a (bin in PATH before ucb)
PID TTY TIME CMD
6791 pts/5 0:00 tcsh
7185 pts/4 0:00 less
7190 pts/5 0:00 ps
% ps -a (ucb before bin)
PID TT S TIME COMMAND
3933 console S 0:00 /usr/lib/saf/ttymon -g -d /dev/console -l console -T xt
6730 pts/1S 0:00 tcsh
6710 pts/2S 0:00 tcsh
6687 pts/3S 0:00 tcsh
6711 pts/4S 0:00 tcsh
7185 pts/4S 0:00 /opt/bin/less usr/src/cmd/ps/ps.c
6686 pts/5S 0:00 tcsh
6791 pts/5T 0:00 /usr/bin/tcsh
7192 pts/5O 0:00 ps -a
Mostly it works nice except, possibly, when you want to get the usage
message.
Casper
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] strange ps behaviour
>I noticed that : ># /bin/ps --help >/bin/ps: illegal option -- help >usage: ps [ -aAdefHlcjLPyZ ] [ -o format ] [ -t termlist ] >[ -u userlist ] [ -U userlist ] [ -G grouplist ] >[ -p proclist ] [ -g pgrplist ] [ -s sidlist ] [ -z zonelist ] [-h > lgrplist] > 'format' is one or more of: >user ruser group rgroup uid ruid gid rgid pid ppid pgid sid taskid > ctid >pri opri pcpu pmem vsz rss osz nice class time etime stime zone > zoneid >f s c lwp nlwp psr tty addr wchan fname comm args projid project > pset lgrp > >whereas : ># /bin/ps help >/bin/ps: illegal option -- h >/bin/ps: illegal option -- p >usage: ps [ -aceglnrSuUvwx ] [ -t term ] [ num ] > >It does feel a bit strange that "usage:" output should be so different, >doesn't it ? Especially when different options are listed > >BTW I noticed that in b132 processes running in zones are listed by "ps ax", I >wasn't aware of tha t, nor can I remember having seen this before, has this changed ? "ps" tries to either pretend to be a "UCB ps" or a "SVR4 ps". --help gets interpreted as "SVR4" and "help" is interpreted as "UCB" (UCB ps didn't required a "-" before all the option types. Nothing has changed in the "ps ax" output. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Solaris vs OpenSolaris (NOW: Read The ARCHIVES)
>On Wed, 03 Feb 2010 21:33:27 -0500, Edward Ned Harvey wrote: > >>> PLEASE GO READ THE MAILING LIST ARCHIVES. >>> >>> This question has been asked, discussed, and answered by several folks >>> (myself included) in fairly exhaustive detail, a large number of times. >>> Most recently, I think, LAST WEEK. >>> >>> Google - use it. >> >> Big deal. I say continue talking about anything you want to talk about. >> It's a free country. I am sure, just as you can find the answer to this > >That depends on which country you are in .. :-) I think it's fine to ask any question; I don't think it is fruitful to tell people to use "google" to find the answer. Unless, of course, google is now sponsoring people to post such remarks; then it is fruitful for the person who tells you to use google? Are you in the pay of Big Oil^H^H^H^H^H^H^HGoogle? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Need to find files with in name. How?
>So I have lots of files with weird characters in the name. Those , >, etc characters are listed as "?", for instance >"Making?Food?Recipe.doc". The version of CIFS I use, (b125) does not >allow listing of files with those characters in WinXP clients. I must >stay with b125 because of issues in later builds. >So I need to somehow, find all these files and rename them in an >automatic way. I have tried ls -R | grep "?" >and then I find all file names with strange characters. But, I dont see >the path to the file. Some where, in all subdirectories there are some >files. But in which sub directory? Where? So I searched >each file and renamed it manually. But that sucks. Hey, I am on Unix, >not on Windows, there should be a way to automate it! Use "find" and list all the filenames: find . then rename them to UTF-8 version of the same characters. (I was happy and I only had two files which weren't displayed) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Asterisk
>Mtmalloc Linked? Calls per secondContext SwitchesUser >CPU % System CPU % >no 90017,000-19,000 66 > 32 >yes 1,40014,000-15,000 66 > 28 > Was it also linked against libumem? I don't think mtmalloc() had really stellar performance. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Atom based board, and experiences?
>Hi Matthias, > >have you also evaluated going the AMD route? I am also looking into >setting up a home server, and I have seen that with AMD one can get ECC >RAM for much less money. Which systems are you thinking of? >Additionally, there seem to be not much difference power consumption >wise when compared to ATOM based setups, when one chooses the right >components. I am a little bit hesitant to go this route, as driver >support for Intel based components seems to be better... The older Atom boards use around 40W, the newer around 20W. Neither have ECC. I certainly would like to hear of a 20W or thereabouts AMD board. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Atom based board, and experiences?
>Super Micro has an attractive board for low power home systems: > > > http://www.supermicro.com/products/motherboard/ATOM/ICH9/X7SPA.cfm?typ=3DH > >Has anyone already put his hands on that thing, or knows, if the actual >OpenSolaris build will/does support the compleet hardware? (Aka: Network, >Graphics, Storage, plus "chipset")? I'm not sure about the ICH9R: it needs to work as a standard AHCI card not in RAID mode. >On a different sidenote: > >Looking at > >http://ark.intel.com/ProductCollection.aspx?familyID=3D29035 > >Intel is offering some very low power Atoms, that also have virtualization in >them (VT-x), anyone already seen board for/with them, and knows, if >OpenSolaris will work? Unfortunately, the Z5xx ATOM CPUs are all 32bit which, I think, disqualifies them. I have a Atom 230 (with the power hungry motherboard) and the new motherboard is clearly better when it comes to power (D410/D510) but I'm not sure it is much faster. The Supermicro boards all use SO DIMMs and I'm not such a fan of those. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] What's Going to Happen to OpenSolaris?
>> SMF is already filled up with so much stuff that at a >> certain point, it starts to become overwhelming from >> a sysadmin persective just to parse through the >> output of svcs -a and now the goal is to add even >> more stuff to it by getting rid of scripting during >> package installation and offloading all of that stuff >> to SMF as well! Jeepers creepers this is going to be >> tough to admin! >> >> SysV packages had shell scripts in them, or at least >> I think they did, but then again I'm no UNIX expert. > > this may help >http://docs.sun.com/app/docs/doc/817-1985/hbrunlevels-25516?a=view For some reason, "package scripts" were apparently "very bad". Unfortunately, it seems that the cure is worse. There are two parts which I really detest: - all the additional scripts are run at boot; they do take time and much more than you realize because they are all run with a cold cache (i.e., all the files needed needed to be found and loaded; a few of the "startup-scripts-only-needed-after-a-package was-update" run find. - "svcs" is now nearly useless: you can't see the forest for the trees I think this clearly needs some architecture applied to it. There are some issues with IPS also; it takes a lot of memory (500 to 1500MBs) and it's a lot slower than SVr4 package install. (On an Ultra 40 with 5GB, it takes about 80 minutes to install "redistributable" from a local package server. That's more time needed to install SXCE and the latter installs more software [25 minutes]. (Yes, I do use "redistributable" having to continuously add extra packages cost more of my time than the local bandwidth: I believe this is a fair comparison; both were installed from a local server over a 100Mbit net and install a similar amount of software) While "memory is cheap" it makes IPS not scale to multiple concurrent zones and makes it impossible to fit on small memory systems (2GB seems to be the minimal if you want to install redistributable) Clearly, serious work needs to be done and I suppose it is good that everyone in the Solaris engineering community is now forced to use IPS. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Goodbye Sun.com
>Well, it's a good thing that for marketing purposes, Sun starting calling >SunOS "Solaris", startin g with SunOS 5. If they hadn't done that, we'd have to to call Solaris "OracleOS" now! No, really, there was a Solaris 1.0.1 (SunOS 4.1.4) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] UltraSPAR_K_
>Cyril Plisko wrote: >> http://www.oracle.com/us/products/servers-storage/servers/blades/index.html >> >> Oh well... >> > >I don't see it. Maybe it's been fixed already? It looked at it again, saw "SPARK"; refreshed it and now it says SPARC. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] OpenSolaris snv_130 panic.
>Experiencing very strange behaviour with upgraded OpenSolaris machine ( from >snv_124 to snv_130 ) > >Machine gets panic, with the following error: > >trap: Unknown trap type 8 in user mode. > >panic[cpu0]/thread = fbc2e3a0: mutex_enter: bad mutex, lp=20 >owner=f000e987f000fea0 thread =fbc2e3a0 > >unix: mutex_panic + 73 () >unix: mutex_vector_enter +446() >genunix: zone_getspecific+2b () >genunix: core+5f () >unix: kern_gpfault+18e () >unix: trap+41e () >unix: cmntrap + e6 () > > >Maybe somebody can help with advice and/or experience the same behaviour. >Machine is IBM xSeries 3650 64 bit. >Running smoothly using snv_124 build. >I've also upgraded the same type of machine from snv_124 to snv_130 just about >day ago with no pro blems. Do you get panics often? Do they all have the same or a similar traceback? (A process dumps core and then the system panics with a bad mutex_ Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] devfsadm:driver failed to attach:
>Hi Casper, > >I placed the files in the directories according to your instructions. > >I got the following messages in /var/adm/messages when I try to do > >#add_drv dummy > >-- >Jan 4 20:58:39 opensolaris genunix: [ID 720415 kern.notice] file >/usr/kernel/drv/amd64/dummy: >Jan 4 20:58:39 opensolaris genunix: [ID 370954 kern.notice] symbol >dummy_dip: >Jan 4 20:58:39 opensolaris genunix: [ID 780480 kern.notice] value >0xc02e1ff0 does not fit >Jan 4 20:58:39 opensolaris genunix: [ID 399259 kern.notice] do_relocations: >/usr/kernel/drv/amd64/dummy do_relocate failed >Jan 4 20:58:39 opensolaris genunix: [ID 603676 kern.notice] dummy error >doing relocations >-- - Looks like progress; you need to use the appropriate compiler and link flags to create a kernel module. Specifically, I think you need -xmodel=kernel. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] devfsadm:driver failed to attach:
>Hi Brian, > >I tried as you said, I placed the actual file 'dummy' in /usr/kernel/dummy >and the conf file 'dummy.conf' also in the same folder. > >But then also its not working. >It asks me to place the driver file (dummy) in correct directory. and to >reboot. > The device driver must be in /usr/kernel/drv; not directly under /usr/ kernel. You should install the files as follows: /usr/kernel/drv/dummy - the binary device driver (32 bit) /usr/kernel/drv/amd64/dummy - the binary device driver (64 bit) /usr/kernel/drv/dummy.conf Assuming x86; with SPARC you'd only create a 64 bit driver in /usr/kernel/drv/sparcv9/dummy Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] snv_130
>casper@sun.com wrote: >> I've noticed that more and more things have stopped working in SXCE; >> specifically X which is now only partially delivered (no Xnest) and, > >We've always only delivered the Xnest built on Xsun, so it's there on >SPARC, been gone on x86 since we removed Xsun on x86 in snv_118. > >Xnest hasn't really been maintained in years, and is missing support >for many of the newer extensions required by modern desktops, which is >why we choose to EOL it in favor of the newer and more capable Xephyr, >built from the current Xorg sources. Sigh, why do people continuously reinvent tools and rename them? It just doesn't make sense. (And, unfortunately, I would think that could be Xnest lot faster than Xephyr.) DESCRIPTION Xephyr is a kdrive server that outputs to a window on a pre-existing "host" X display. Think Xnest but with support for modern extensions like composite, damage and randr. "kdrive server"?? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] snv_130
>Happy new year every1! > > >> 130 is the final SXCE release planned though, as previously announc= >ed. > >Which is too bad, really. Maybe the infrastructure to build SXCE >can be kept in place a little bit longer. The final release of SXCE >could then be cut in sync with the official OS 2010.02 release. I've noticed that more and more things have stopped working in SXCE; specifically X which is now only partially delivered (no Xnest) and, unfortunately, GDM was upgraded so getting Sun Ray to work is impossible, it seems, except by using dtlogin. The time you need to install OpenSolaris is much larger than installing SXCE, mostly trying to find which packages you've missed this time. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] How to cheat Sun server that cover is closed?
>Hi, > >I have a Sun Sparc T5240 server. I want to open the cover to hook my test >instruments. But, if the cover is open, then the server won't boot up. > >How do I cheat the server that cover is always closed? If you run the system with an open cover, you will disturb the airflow and the system may die. That's why there's a switch. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Syslog-ng?
>> Solaris Syslog doesn't log using a door. >> >> Solaris logs on "/dev/conslog" but uses the door only >> to determine if the >> syslogd is running. >> >> Casper > >I always wondered at the sanity of using a door for this purpose. My biggest problem with the code is that syslogd_ok() is always called, even when we're not using the data. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Syslog-ng?
>On Mon, Dec 7, 2009 at 8:34 AM, Eugene Vilensky = > wrote: >> Hi, >> Is syslog-ng available for OpenSolaris, through IPS or otherwise? = >=A0If >> I were to build it, is there an "easy" metapackage for the build >> system, maybe analogous to Red Hat's rpmbuild+dependencies or Debia= >n's >> "build-essential" metapackage? > >I doubt that if it were installed on OpenSolaris that it would work >for processing logs for local programs. This is because the things >that do the logging are going to use libc's implementation of >syslog(3C), which logs via a door (see libdoor(3lib)). Doors are >unique to [Open]Solaris and as such syslog-ng is unlikely to be >listening for log messages at /etc/.syslog_door. Solaris Syslog doesn't log using a door. Solaris logs on "/dev/conslog" but uses the door only to determine if the syslogd is running. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] lx zone no /dev/mem or /dev/kmem
>Yes Casper. This is not open source application. >Rather it is Server administration application and while installing, it >checks for server id >to check whether installation is done on supported server system. >My doubt is that server ids are stored in predefine memory location of >particular server and >installer script tries to read those memory using /dev/mem. > >Do you have any input on this? Without knowing exactly what it does, it is difficult to "fake" this. But typically, when systems look into kernel memory they have a list of kernel symbols, a path to open the kernel symbol list and the pathname to the memory. The application will use nlist to translate the names to addresses and then it will seek to /dev/mem. Not too difficult to fake that but I'm afraid that is what you need to do. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] exec_attr is empty..
>I've tried to reproduce it. Unsucessfully. Everything updated normally. Maybe >because this time pa ckage was taken from cache... Can't happen when you install a SVr4 package; if exec_attr is empty, then it is either: a SVr4 package with a non-standard i.rbac install script or IPS bug (which is what you encontered probably) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] How to make syslogger write to a fifo?
>Will keep reading so any ouput from system logger is displayed... but
>a regular kind of command like:
>
> cat named-pipe
>
>cat will close after the first line is read.
Not my experience; "cat" works.
>Similar for something like awk '/REGEX/{print}' named-pipe > some.log
>
>awk will close after the first line of input.
>
>I guess I need to write a script that behaves like `tail -f' and I'm
>sure I can in perl, but is that really necessary?
>Isn't there some way to get a standard filtering script to get all
>lines the system logger puts out, without trickery inside the script
>to keep it from closing after the first line.
>
>I guess the linux way of using a pipe `|' in the actual syslog.conf is
>one way to get it done... but that does not work on opensol.
Well, you have the source (though I understand that your code skill
are light)
Casper
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] How to make syslogger write to a fifo?
>Can anyone tell me how to make the system logger write to a fifo? > >I've tried this in /etc/syslog.conf > > *.debug |/var/adm/flt > >Where `flt' is a fifo and the two elements are separated by tabs. I'm assuming that "flt" is a named pipe; doesn't it work when you use /var/adm/flt? I just tested this and it does seems to work. But syslog will close it when nothing is reading from the pipe. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] ulimit -u not working
>I understand, but I was just pointing out that the syntax was not wrong. >I googled a bit on it and only found people have problems setting it to a >higher value. >like 5 and how one should do it. There's no such settable limit in Solaris. While bash/ksh93 print something on "ulimit -u" it's not settable and it's not a resource that is easily changable. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] doubt in solaris
>can we run objective-c programming language in solaris,because it will be run >only in Mac os You will need an Objective-C compiler. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] limit number of file-descriptors in system.
>Hi Casper. > >I'd like to run hundreds process, and they have thousands file-descriptors. >Therefore, I'd like to limit them in system. So that processes will fail? What problem are you trying to solve? >Sorry, I don't know max number of them in system level. There is no such limit so you cannot change the limit to a lower value. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] limit number of file-descriptors in system.
>Hi, all. > >I'd like to limit number of file-descriptors in system. >What do I set a parameter? > >In process level, I know "rlim_fd_max" paramter. There's no system wide limit. Why do you want this? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Point-of-View ION+Atom 330
>Hi, > >I have recently bought this motherboard and I was plaining use >OpenSolaris so I went and installed release 2009/06 but it seems >the release doesn't support video/network card. > >http://www.pointofview-online.com/showroom.php?shop_mode=product_detail&product_id=97 > >So I installed svn_123b instead so the network card driver is working >but the video isn't, also so wrong "tpm failed to attach", >is there any plan to support nVida ION motherboards. Tpm is a bug: you'll find it on any system w/o tpm installed. (6874163) I'm not sure about the ION; whether it will be supported by the standard nVidia driver. Their website seems to indicate that it does: http://www.nvidia.com/object/solaris_display_185.18.36.html Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [ug-bosug] usb 2.0 works slow
>> i was trying to achieve the speeds mentioned according to the usb 2.0 specs. >> http://en.wikipedia.org/wiki/USB_2.0#USB_2.0 > >Hi Arun, > >What is the speed that you are expecting? >Line speed for USB Hi speed is 480 Mbits/s. >So it is 60 MB/s. >(Note the differences ib M*Bits* and M*Bytes*) > >You are getting : >~18 MB/s on Read >~4 MB/s on Write >> a...@opensolaris:/media# rmformat >> Looking for devices... >> 1. Logical Node: /dev/rdsk/c11t0d0p0 >> Physical Node: /p...@0,0/pci1043,8...@13,5/stor...@6/d...@0,0 >> Connected Device: Kingston DT 101 II1.00 >> Device Type: Removable >> Bus: USB >> Size: 15.3 GB >> Label: repo0906 >> Access permissions: Medium is not write protected. Kingston say the maximum read/write speed is 10MB/s (both ways). Not sure why the "Read" returns more quickly than the device specifies. (http://www.kingston.com/flash/dt_chart.asp) Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] External disk >1TB
>I tried also with format (on an Opensolaris 64-bit where I can see th= >e >disk), then connected the disk back on the 32-bit system, same error. > >Any other ideas? Disks > 1TB will just not work on a 32 bit Solaris. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] What happened to rarpd?
>All; > >This weekend I installed osol 0906 on my laptop and in the process of >copying >files and testing configurations found that there is no rarpd. I need >rarpd for >performing Jumpstart/JET builds on SPARC based systems. > >How can I get this installed on my laptop asap? Why not using DHCP (boot net:dhcp - install) for SPARC? Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] How to delete files as dos 'del *.java /s' on command line
>thank you,
>
>can you tell me what does it mean {} + ?
>can you suggest me a tutorial or an online manual for beginners?
>--
"man find".
{} -> expanded to the name of the files found
+ -> call the command with multiple file arguments
Casper
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Nokia Booklet 3G
>As far as I know there are closed binaries drivers >produced by hardware vendors for Linux world. To be >able to run such drivers on Linux there is a piece of >code licensed under LGPL allowing such thing. That is >the purpose of LGPL. Some hardware vendors still do >not want to write open source drivers because they'd >uncover their architecture to the competitors. You can ship closed drivers for Linux but not as part of your product; i.e., you can ship you own driver and open only the "glue" which connects to the Linux kernel. But if you make a Linux distribution, you cannot ship such drivers but users can download and deploy them. Also, when you ship a device which runs Linux you are required to open your kernel but not necessarily all the applications which are part of your offering. Casper ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
