[9fans] unvac troubles...
Hi, I'm using unvac to try to extract some vac archives I made in the fall of 2008. I'm running into a pair of problems. First, unvac is outputting directories with the write bit off. This causes it to fail pretty early, as it can't write files to the newly-extracted directories. Is this expected? Is there a way to avoid this problem? Second I get this error: (on p9p unvac) unvac: pthread_mutex_lock.c:289: __pthread_mutex_lock: Assertion `(-(e)) != 3 || !robust' failed. [New Thread 0xb7e548d0 (LWP 3743)] Program received signal SIGABRT, Aborted. [Switching to Thread 0xb7e548d0 (LWP 3743)] 0xb8022430 in __kernel_vsyscall () (gdb) bt #0 0xb8022430 in __kernel_vsyscall () #1 0xb7fc64b0 in raise () from /lib/tls/i686/cmov/libpthread.so.0 #2 0x0805c6b2 in child () at daemonize.c:38 #3 0x0805c927 in _threadsetupdaemonize () at daemonize.c:153 #4 0x0805c49d in p9main (argc=3, argv=0xbfa3e1c4) at thread.c:714 #5 0x08060d92 in main (argc=Cannot access memory at address 0xe9f ) at main.c:10 (gdb) Have any of you seen anything like this? -- vs
Re: [9fans] pipefrom fun
why not use /mail/lib/names.local for the first example? good point, upon my brief look the answer would be, although I can contrive upon boot to use a different file, in the general case, users can't edit it. i haven't read the script, though. ; ls -l /n/sources/contrib/maht/rc/pipefrom --rwx--x--x M 675 maht sys 1107 Aug 20 20:50 /n/sources/contrib/maht/rc/pipefrom I fixed that, oops thanks one thing i've thought would be useful since people mail from many places with many names is a translation from a set of from addresses to a mailbox name so that, e.g. the ned f command doesn't end up creating a bunch of folders for the same person. I'll consider that when I get round to pipeto
Re: [9fans] unvac troubles...
On Fri, Aug 21, 2009 at 2:15 PM, Venkatesh Srinivasm...@acm.jhu.edu wrote: Hi, I'm using unvac to try to extract some vac archives I made in the fall of 2008. I'm running into a pair of problems. First, unvac is outputting directories with the write bit off. This causes it to fail pretty early, as it can't write files to the newly-extracted directories. Is this expected? Is there a way to avoid this problem? This happens to me too. I usually just delete the old directory and extract again. fernan -- http://www.fernski.com
Re: [9fans] pipefrom fun
Russ Cox wrote: pipefrom is too late to rewrite outgoing addresses, because marshal has already written the To: I submitted a patch for marshal to do the From: part but was told I should use pipefrom instead. I'll stick a sed in the send see how that works out : line with the local address. Your post says To: 9f...@steponnopets.net dammit, I obviously had missed that when I was testing it You can create /mail/box/$user/names to get aliases that won't go out. http://9fans.net/archive/2004/01/530 http://9fans.net/archive/2004/01/533 http://9fans.net/archive/2004/04/99 # ,s/%/@/g I think Thanks I'll give those a read
[9fans] venti/copy -m
Hi, Plan 9's venti/copy has an undocumented -m option. What does it do? Thanks, -- vs
Re: [9fans] venti/copy -m
Plan 9's venti/copy has an undocumented -m option. What does it do? the whole program is 262 lines long. i'm betting what -m does can be discovered by inspection. it might be a good idea to submit a patch to the man page, too. - erik
Re: [9fans] pipefrom fun
On Fri, Aug 21, 2009 at 12:59 AM, mattmaht-9f...@maht0x0r.net wrote: pipefrom is too late to rewrite outgoing addresses, because marshal has already written the To: I submitted a patch for marshal to do the From: part but was told I should use pipefrom instead. I'll stick a sed in the send see how that works out : They're separate issues, which might not have come across in my earlier mail. Rewriting of To: must be done in marshal and is supported by editing /mail/box/$user/names. Rewriting of From: is the reason for pipefrom. Russ
[9fans] Plan 9 via QEMU
... on Acer Aspire One to add more fun. Yes, on the Acer Aspire One, I have no means to compile, with all hopes of GCC thwarted by the draconian schemes of archaic Linux package managers, all slowly drowning in dependency hell. But enough fantasising -- No GCC, no kernel sources (which are presumably highly customised for the hardware), so no extra kernel modules. This means bridging as well as filtering are out the window, since I lack bridge.ko, tun.ko, and ipt_wtf.ko (though interestingly enough I found tun.ko that someone had precompiled for the AAO). And yet, I need to get QEMU to PXE boot Plan 9 and then use the local Ken FS server for root filesystem. The key was a different style of networking here - the pcap[1] device. It uses the host system's device directly, so if I want to PXE boot Plan 9 inside QEMU, I have to 'ifconfig ath0 0.0.0.0' first, but it works. The way around that would be to use a different MAC address for the emulated device, which is possible with plain ether- net NIC, and with certain wireless config- urations too, but ath0 won't allow it for some reason. So, QEMU starts up[2], PXE boots Plan 9, which gets to the root is from (tcp, il) [il]: prompt just fine, and subsequently the user and password queries - all of this goes by fine, and indicates network connectivity with the CPU/auth server, but suddenly everything kernel panics, I presume at the time of making the connection with the Ken FS server (scrolls by too fast for me to see the initial error message). If anyone has any experience or ideas regarding that, I'd love to hear it. Best, ak P.S.: I had to statically compile QEMU with the pcap patch[1] on a mate's machine, long distance, and copy the binary back. Version 0.9.1. No SDL, for it's a bitch to statically compile, thus have to use VNC to see output. [1] http://qemu-forum.ipi.fi/viewtopic.php?f=2t=1568 pcap patch for those curious or interested [2] QEMU startup command: sudo ./qemu -L bios -option-rom bios/pxe-rtl8139.bin \ -boot n -net nic,model=rtl8139,macaddr=00:22:68:C3:4C:08 \ -net pcap,devicename=ath0 -no-kqemu -soundhw sb16 \ -localtime -no-acpi -vnc :1 (yes, sudo: this is Linux, we have rules... and a dictatorship to defeat their purpose)
Re: [9fans] Recursive structural expressions?
On Aug 20, 2009, at 21:01, erik quanstrom quans...@quanstro.net wrote:
Here's an example. Let's make the syntax extra pukey: @#, where # is
1-9, defines a `named procedure', which is the same thing as putting
something in braces in Sam.
x/.*\n/ @1{ ( @1 ) | @1 ( @1 ) ( ) | }
x/re/ repeatedly sets . with matches until the input
is exhausted. so i don't think i understand your example.
how does /.*\n/{(.*\n)|.*\n(.*\n)|} match anything
but the tautological .*\n|? for example, if the input
is line1\nline2\n, then . is in turn line1\n and line2\n,
and both match .*\n|.
where do the actions go in your example?
I guess I botched the example a bit. The whole part after the @ should
be in an x//, and the actions would come after. Something more
complicated would allow you to match any parenthetical expression on a
line and alter it, which could be useful.
- erik
Re: [9fans] venti/copy -m
On Fri, Aug 21, 2009 at 6:15 AM, Venkatesh Srinivasm...@acm.jhu.edu wrote: Plan 9's venti/copy has an undocumented -m option. What does it do? On Fri, Aug 21, 2009 at 6:23 AM, erik quanstromquans...@quanstro.net wrote: the whole program is 262 lines long. i'm betting what -m does can be discovered by inspection. what, maybe, but not always why. it's a valid question. it might be a good idea to submit a patch to the man page, too. replace The .B -f option causes .I copy to run in `fast' mode, assuming that if a block already exists on the destination Venti server, all its children also exist and need not be checked. with Venti's blocks are arranged in a directed acyclic graph (see venti(6)); there may be multiple paths from a root score to an interior block (for example, if the same file contents are stored under multiple names in an archive). .I Copy runs more efficiently if it does not copy blocks (and all their children) multiple times. The .B -f option causes .I copy to assume that if a block already exists on the destination Venti server, all its children also exist and need not be considered. The .B -m option causes .I copy to maintain an in-memory list of blocks it has copied and avoid considering the same block multiple times. The .B -f option is only useful if the destination Venti server is known not to have lost any blocks due to disk corruption or other failures. The .B -m option is only useful enough memory is available to hold the block list, which typically requires about 1% of the total number of bytes being copied.
Re: [9fans] Recursive structural expressions?
Regular expressions, in the strict formal sense, have an important property: they completely express the set of patterns that can be searched for in a single linear-time pass through the text. That is, they have an associated linear-time performance guarantee. In your particular case, adding recursion would produce context-free expressions, and general context-free parsing as typically implemented is cubic as opposed to linear. (It turns out to be the same problem as matrix multiplication, so with a lot more effort you could get it down to n^2.38 or perhaps even lower, but you can't possibly get sub-quadratic.) As Charles pointed out, there is a long, distinguished history of editors and other tools making regular expressions more expressive. It is important to remember that this is always done at the cost of giving up the performance guarantee, at least in some cases. And unless a general matcher can distinguish the different cases and implement separate algorithms for each, it usually gives up the performance guarantee in all cases. Russ
[9fans] venti/copy question....
Hi, I have a few score trees in Venti that p9p's venti/copy wouldn't copy, but p9's would. venti/copy aborted with: copy: reading block (type 16): read asked for got da39a3ee5e6b4b0d3255bfef95601890afd80709 P9's venti/copy has a test, in its walk() function, as to whether a score is all-zeros. P9P's venti/copy doesn't. Adding this test allows my copy to proceed. Is there any reason p9's venti/copy has this test? It seems like a dubious one, since the all-zero score shouldn't ever be in Venti. But it's also necessary to copy these trees... Thanks, -- vs
Re: [9fans] Plan 9 via QEMU
you can configure Qemu to use the host console as the guest serial console
Re: [9fans] venti/copy -m
Thanks for the answer, In p9's venti/copy, in scoretreecmp, there are two casts, from Avl* to ScoreTree*; this depends on scoretree's avl being the first member of the structure; I thought kencc was allowed to reorder structure members, is that the case? Thanks, -- vs
Re: [9fans] pipefrom fun
Rewriting of From: is the reason for pipefrom. in the simple case, From: can be rewritten with /mail/box/$upasname/headers - erik
Re: [9fans] venti/copy -m
I thought kencc was allowed to reorder structure members, is that the case? No.
Re: [9fans] Recursive structural expressions?
On Aug 21, 2009, at 10:24, Russ Cox r...@swtch.com wrote: Regular expressions, in the strict formal sense, have an important property: they completely express the set of patterns that can be searched for in a single linear-time pass through the text. That is, they have an associated linear-time performance guarantee. In your particular case, adding recursion would produce context-free expressions, and general context-free parsing as typically implemented is cubic as opposed to linear. (It turns out to be the same problem as matrix multiplication, so with a lot more effort you could get it down to n^2.38 or perhaps even lower, but you can't possibly get sub-quadratic.) As Charles pointed out, there is a long, distinguished history of editors and other tools making regular expressions more expressive. It is important to remember that this is always done at the cost of giving up the performance guarantee, at least in some cases. And unless a general matcher can distinguish the different cases and implement separate algorithms for each, it usually gives up the performance guarantee in all cases. Russ This, and complexity/ambiguity are almost certainly why they didn't carry from qed into ed. Perhaps at least a context-free grep would be a useful thing to have, considering the limitations CF expressions would place on a text editor.
Re: [9fans] Plan 9 via QEMU
Upon Steve's suggestion, I enabled *nodumpstack=1 in the PXE config for the Plan 9 in QEMU. The result follows: ... password: ! time... panic: boot process died: sys: demand load I/O error accessing /386/init: mount rpc error panic: boot process died: sys: demand load I/O error accessing /386/init: mount rpc error dumpstack disabled cpu0: exiting So, could it be that there are problems doing IL through QEMU? There shouldn't be, as the pcap device sends directly through the interface... so far as I can tell. Input welcome. Thanks, ak
Re: [9fans] Plan 9 via QEMU
So, could it be that there are problems doing IL through QEMU? There shouldn't be, as the pcap device sends directly through the interface... so far as I can tell. Input welcome. do a packet trace from the host. it's also possible to turn on some il tracing on the fs. here are a couple of the flags from my fs that might be interesting: ../dev/cw.c:401:roflag = flag_install(ro, -- ro reads and writes); ../ip/arp.c:72: arpcache.flag = flag_install(arp, -- verbose); ../ip/il.c:170: ilflag = flag_install(il, -- on errors); ../port/con.c:830: attachflag = flag_install(attach, -- attach calls); ../port/con.c:831: chatflag = flag_install(chat, -- verbose); ../port/con.c:832: allowflag = flag_install(allow, -- allow mode); ../port/con.c:834: whoflag = flag_install(allchans, -- on who); ../port/con.c:835: authdebugflag = flag_install(authdebug, -- report authentications); ../port/con.c:836: authdisableflag = flag_install(authdisable, -- disable authentication); (i believe in the standard 64bit fs, allow isn't a flag. i did it that way, so i could allow one channel.) the syntax is flag $flag. it's a toggle. you can also apply a flag to a single channel, but i don't think you've gotten that far. flag $chan $flag. - erik
Re: [9fans] venti/copy question....
thanks for the tip. i got the same error (even the score, i think) when i tried copying some vbackup scores to a new server a few weeks ago. i hadn't thought of trying a different copy until you brought it up. using plan 9's copy instead of p9p's seems to be working well enough so far
[9fans] Issues with 2 networks, fs server, and namespaces
Hello all, I'm trying to set up a group of servers (these are running on VMWare ESXi, and working great -- CPU server running with two APs, though adding more causes it to fault with a divide by zero?). Auth server's got its own 1GB fossil, boots with the 9pcauth kernel. CPU server boots from a small fossil. Both Auth and CPU are on the public internet via ether0 so that they are cpu/drawtermable. They do not boot from the file server because I didn't want to set up a DHCP server that was connected to the Internet (ISP getting mad and whatnot). While I've configured the internal network to be on it's own vswitch (managed through vmware, no real network connectivity), I've been struggling with the prior configuration enough that I don't want to just `give up' on it. The FS, however, sits on a private network. CPU and Auth are connected to this network via ether1. However, I'm having the following issues: #1) Using two networks on two different interfaces is a pain in the ass. I've got: bind '#l1' /net.alt bind '#I1' /net.alt in my /cfg/cpu/namespace. If I simply have them here, ip/ipconfig -N -x ether1 ether /net.alt/ether1 complains in cpurc about no ip being attached to /net.alt. So I have to put that in /cfg/cpu/cpurc also. I don't quite understand why everything's architected to have a single ip stack on a single ethernet; in this case, it really isn't convenient that it doesn't determine the correct interface via routing tables or somesuch. Is there something basic that I'm missing here? #2) Drawterm is taking forever and a day to connect and log in. It's either an auth issue or a DNS issue. Best guesses as to what this could be and how I should go about diagnosing it? #3) Trying to mount the fileserver globally is elusive. I want to mount /n/fs/usr over /usr and /n/fs/mail over /mail. Perfectly happy with that. However: o Doing that in cpurc doesn't put it in the global namespace o Doing it in /cfg/cpu/namespace doesn't have an ip yet so I can't run srv /net.alt/tcp!10.0.0.3!9fs in the first place o Doing it in /rc/bin/service/tcp17010 causes me to get `cpu: negotiating authentication method: [public auth server ip]: cs gave empty translation list' Mounting it from /n/fs after booting works fine (but it makes me auth, which is kind of weird -- I guess I need to set up a secstore? -- I figured that eve would be able to connect without auth, given that everything's tied to the same auth server, no matter which network it's on, and that a user drawterming in would be able to connect by virtue of having authed when connecting in the first place.) I know the `preferred way' is to boot the CPU server from the fileserver. While I could feasibly reconfigure my setup to do this, I'd prefer to figure it out this way first, given the amount of time I've been banging my head against the wall on it :) --dho
Re: [9fans] Issues with 2 networks, fs server, and namespaces
You don't need a second IP stack. You can run both interfaces on the same IP stack and routing will just work. That's how I did it when I had a similar setup. -Chris On Fri, Aug 21, 2009 at 14:07, Devon H. O'Delldevon.od...@gmail.com wrote: Hello all, I'm trying to set up a group of servers (these are running on VMWare ESXi, and working great -- CPU server running with two APs, though adding more causes it to fault with a divide by zero?). Auth server's got its own 1GB fossil, boots with the 9pcauth kernel. CPU server boots from a small fossil. Both Auth and CPU are on the public internet via ether0 so that they are cpu/drawtermable. They do not boot from the file server because I didn't want to set up a DHCP server that was connected to the Internet (ISP getting mad and whatnot). While I've configured the internal network to be on it's own vswitch (managed through vmware, no real network connectivity), I've been struggling with the prior configuration enough that I don't want to just `give up' on it. The FS, however, sits on a private network. CPU and Auth are connected to this network via ether1. However, I'm having the following issues: #1) Using two networks on two different interfaces is a pain in the ass. I've got: bind '#l1' /net.alt bind '#I1' /net.alt in my /cfg/cpu/namespace. If I simply have them here, ip/ipconfig -N -x ether1 ether /net.alt/ether1 complains in cpurc about no ip being attached to /net.alt. So I have to put that in /cfg/cpu/cpurc also. I don't quite understand why everything's architected to have a single ip stack on a single ethernet; in this case, it really isn't convenient that it doesn't determine the correct interface via routing tables or somesuch. Is there something basic that I'm missing here? #2) Drawterm is taking forever and a day to connect and log in. It's either an auth issue or a DNS issue. Best guesses as to what this could be and how I should go about diagnosing it? #3) Trying to mount the fileserver globally is elusive. I want to mount /n/fs/usr over /usr and /n/fs/mail over /mail. Perfectly happy with that. However: o Doing that in cpurc doesn't put it in the global namespace o Doing it in /cfg/cpu/namespace doesn't have an ip yet so I can't run srv /net.alt/tcp!10.0.0.3!9fs in the first place o Doing it in /rc/bin/service/tcp17010 causes me to get `cpu: negotiating authentication method: [public auth server ip]: cs gave empty translation list' Mounting it from /n/fs after booting works fine (but it makes me auth, which is kind of weird -- I guess I need to set up a secstore? -- I figured that eve would be able to connect without auth, given that everything's tied to the same auth server, no matter which network it's on, and that a user drawterming in would be able to connect by virtue of having authed when connecting in the first place.) I know the `preferred way' is to boot the CPU server from the fileserver. While I could feasibly reconfigure my setup to do this, I'd prefer to figure it out this way first, given the amount of time I've been banging my head against the wall on it :) --dho -- Christopher Nielsen They who can give up essential liberty for temporary safety, deserve neither liberty nor safety. --Benjamin Franklin
Re: [9fans] Issues with 2 networks, fs server, and namespaces
2009/8/21 Devon H. O'Dell devon.od...@gmail.com: 2009/8/21 Christopher Nielsen cniel...@pobox.com: You don't need a second IP stack. You can run both interfaces on the same IP stack and routing will just work. That's how I did it when I had a similar setup. Wait, I misread your explanation. Would you care to explain more about that? Is that just binding '#l1' into /net? --dho
Re: [9fans] Issues with 2 networks, fs server, and namespaces
Hey Devon, 1. Others know more about that than I do. Wait a bit, that problem might get solved. 2. drawterm tends to hang on secstore for me. Try a bogus -s option or use a p9p secstore/factotum and see what happens. 3. what's stopping you from setting up your external network as the one on /net.alt and using dhcp internally? Or booting tcp off the file server without dhcp? e.g. baking it into plan9.ini, or entering it manually at boot time? I've tried similar tricks to yours using inferno and getting cute and breaking the convention of /net.alt for external networks has always ended in a world of pain. Noah On Fri, Aug 21, 2009 at 11:07 PM, Devon H. O'Delldevon.od...@gmail.com wrote: Hello all, I'm trying to set up a group of servers (these are running on VMWare ESXi, and working great -- CPU server running with two APs, though adding more causes it to fault with a divide by zero?). Auth server's got its own 1GB fossil, boots with the 9pcauth kernel. CPU server boots from a small fossil. Both Auth and CPU are on the public internet via ether0 so that they are cpu/drawtermable. They do not boot from the file server because I didn't want to set up a DHCP server that was connected to the Internet (ISP getting mad and whatnot). While I've configured the internal network to be on it's own vswitch (managed through vmware, no real network connectivity), I've been struggling with the prior configuration enough that I don't want to just `give up' on it. The FS, however, sits on a private network. CPU and Auth are connected to this network via ether1. However, I'm having the following issues: #1) Using two networks on two different interfaces is a pain in the ass. I've got: bind '#l1' /net.alt bind '#I1' /net.alt in my /cfg/cpu/namespace. If I simply have them here, ip/ipconfig -N -x ether1 ether /net.alt/ether1 complains in cpurc about no ip being attached to /net.alt. So I have to put that in /cfg/cpu/cpurc also. I don't quite understand why everything's architected to have a single ip stack on a single ethernet; in this case, it really isn't convenient that it doesn't determine the correct interface via routing tables or somesuch. Is there something basic that I'm missing here? #2) Drawterm is taking forever and a day to connect and log in. It's either an auth issue or a DNS issue. Best guesses as to what this could be and how I should go about diagnosing it? #3) Trying to mount the fileserver globally is elusive. I want to mount /n/fs/usr over /usr and /n/fs/mail over /mail. Perfectly happy with that. However: o Doing that in cpurc doesn't put it in the global namespace o Doing it in /cfg/cpu/namespace doesn't have an ip yet so I can't run srv /net.alt/tcp!10.0.0.3!9fs in the first place o Doing it in /rc/bin/service/tcp17010 causes me to get `cpu: negotiating authentication method: [public auth server ip]: cs gave empty translation list' Mounting it from /n/fs after booting works fine (but it makes me auth, which is kind of weird -- I guess I need to set up a secstore? -- I figured that eve would be able to connect without auth, given that everything's tied to the same auth server, no matter which network it's on, and that a user drawterming in would be able to connect by virtue of having authed when connecting in the first place.) I know the `preferred way' is to boot the CPU server from the fileserver. While I could feasibly reconfigure my setup to do this, I'd prefer to figure it out this way first, given the amount of time I've been banging my head against the wall on it :) --dho
Re: [9fans] Plan 9 via QEMU
This time with just flag authdebug and without going through secstore, since it gave me the problem highlighted above, plain auth from the QEMU host gives the following at Ken FS: il: allocating ... user akumar = 2 authenticated authorize: uid is 2 authorize: uid is 2 hangup! connection timed out-3 ... so at least basic auth is being done -- but what's the bizarre problem when doing secstore (notice it hangs around for a long time before giving that error message in the previous post)? And of course, why the time out from the Ken FS server? I've attached a conversation from start to finish between Ken FS (192.168.1.3) and the Plan 9 inside QEMU (192.168.1.20). tcpdump command was: tcpdump -w kenfs-full.pcap-X -s 0 -n -e \ -ttt -i ath0 src 192.168.1.3 or dst 192.168.1.3 to produce the pcap output, and tcpdump -X -nnr kenfs-full.pcap kenfs-full.convo to produce the attached file with human-readble data. Please let me know if more information is desired. Thanks, ak
Re: [9fans] Issues with 2 networks, fs server, and namespaces
2009/8/21 Noah Evans noah.ev...@gmail.com: Hey Devon, 1. Others know more about that than I do. Wait a bit, that problem might get solved. I think I brought that up because if anybody has ideas about fixing them or making them better, I would like to do that. 2. drawterm tends to hang on secstore for me. Try a bogus -s option or use a p9p secstore/factotum and see what happens. I'm not sure I have secstore set up. Perhaps that's it? 3. what's stopping you from setting up your external network as the one on /net.alt and using dhcp internally? Or booting tcp off the file server without dhcp? e.g. baking it into plan9.ini, or entering it manually at boot time? I've tried similar tricks to yours using inferno and getting cute and breaking the convention of /net.alt for external networks has always ended in a world of pain. Tried to explain that. I just don't want to due to the number of hours I've spent trying to configure it this way :). I think Christopher's suggestion is what I'll do next; I totally forgot that was possible. --dho Noah On Fri, Aug 21, 2009 at 11:07 PM, Devon H. O'Delldevon.od...@gmail.com wrote: Hello all, I'm trying to set up a group of servers (these are running on VMWare ESXi, and working great -- CPU server running with two APs, though adding more causes it to fault with a divide by zero?). Auth server's got its own 1GB fossil, boots with the 9pcauth kernel. CPU server boots from a small fossil. Both Auth and CPU are on the public internet via ether0 so that they are cpu/drawtermable. They do not boot from the file server because I didn't want to set up a DHCP server that was connected to the Internet (ISP getting mad and whatnot). While I've configured the internal network to be on it's own vswitch (managed through vmware, no real network connectivity), I've been struggling with the prior configuration enough that I don't want to just `give up' on it. The FS, however, sits on a private network. CPU and Auth are connected to this network via ether1. However, I'm having the following issues: #1) Using two networks on two different interfaces is a pain in the ass. I've got: bind '#l1' /net.alt bind '#I1' /net.alt in my /cfg/cpu/namespace. If I simply have them here, ip/ipconfig -N -x ether1 ether /net.alt/ether1 complains in cpurc about no ip being attached to /net.alt. So I have to put that in /cfg/cpu/cpurc also. I don't quite understand why everything's architected to have a single ip stack on a single ethernet; in this case, it really isn't convenient that it doesn't determine the correct interface via routing tables or somesuch. Is there something basic that I'm missing here? #2) Drawterm is taking forever and a day to connect and log in. It's either an auth issue or a DNS issue. Best guesses as to what this could be and how I should go about diagnosing it? #3) Trying to mount the fileserver globally is elusive. I want to mount /n/fs/usr over /usr and /n/fs/mail over /mail. Perfectly happy with that. However: o Doing that in cpurc doesn't put it in the global namespace o Doing it in /cfg/cpu/namespace doesn't have an ip yet so I can't run srv /net.alt/tcp!10.0.0.3!9fs in the first place o Doing it in /rc/bin/service/tcp17010 causes me to get `cpu: negotiating authentication method: [public auth server ip]: cs gave empty translation list' Mounting it from /n/fs after booting works fine (but it makes me auth, which is kind of weird -- I guess I need to set up a secstore? -- I figured that eve would be able to connect without auth, given that everything's tied to the same auth server, no matter which network it's on, and that a user drawterming in would be able to connect by virtue of having authed when connecting in the first place.) I know the `preferred way' is to boot the CPU server from the fileserver. While I could feasibly reconfigure my setup to do this, I'd prefer to figure it out this way first, given the amount of time I've been banging my head against the wall on it :) --dho
Re: [9fans] Issues with 2 networks, fs server, and namespaces
Well, we're getting somewhere. Using /cfg/cpu/namespace still seems to do nothing to get ether1 into /net. Putting it into cpurc does the trick though, go figure. However, I've got a new issue. When I go to mount the file server, I'm getting this: mount: auth_proxy: auth_proxy rpc write: p9...@int.9vx.org: no key matches proto=p9sk1 dom=int.9vx.org role=client user? !password? mount: mount /n/fs: fossil authCheck: auth protocol not finished I mentioned I wasn't sure that I'd set up secstore. I think I really mean, ``I haven't set up secstore, and I'm not sure how :) Tips? --dho
Re: [9fans] Issues with 2 networks, fs server, and namespaces
And that's taken care of. Didn't have an authdom configured in /lib/ndb/local, and for some reason, I forgot to set up keyfs on the auth server. Thought I had that taken care of. Of course, I'm now faced with another new issue. auth/debug looks like it just tries to debug factotum keys. This machine has an interface on 9vx.org and another on int.9vx.org. However, auth/debug only tries to debug 9vx.org, leading me to believe that factotum has no key for int.9vx.org. Still getting that auth protocol not finished for int.9vx.org, but I'm sure if I could get it to put that key in place, everything would be great. Ideas? --dho
Re: [9fans] Issues with 2 networks, fs server, and namespaces
Of course, I'm now faced with another new issue. auth/debug looks like it just tries to debug factotum keys. This machine has an interface on 9vx.org and another on int.9vx.org. However, auth/debug only tries to debug 9vx.org, leading me to believe that factotum has no key for int.9vx.org. Still getting that auth protocol not finished for int.9vx.org, but I'm sure if I could get it to put that key in place, everything would be great. auth domain != dns domain. there's no reason for them even to be similar. (except human convienence.) - erik
Re: [9fans] Issues with 2 networks, fs server, and namespaces
I'm trying to set up a group of servers (these are running on VMWare ESXi, and working great -- CPU server running with two APs, though adding more causes it to fault with a divide by zero?). Auth server's could you be more specific about this? The FS, however, sits on a private network. CPU and Auth are connected to this network via ether1. However, I'm having the following issues: #1) Using two networks on two different interfaces is a pain in the ass. I've got: bind '#l1' /net.alt bind '#I1' /net.alt you want something like this in /lib/namespace.machinewith2ndipstack cat /lib/namespace.ladd bind -b #l1 /net.alt bind -b #I1 /net.alt mount -a /srv/cs_net.alt /net.alt mount -a /srv/dns_net.alt /net.alt # and you want something like this in your cpurc bind -b '#l1' /net.alt bind -b '#I1' /net.alt ip/ipconfig -x /net.alt -g 192.168.10.254 ether /net.alt/ether1 add 192.168.10.1 /120 ndb/cs -x /net.alt -f /lib/ndb/external ndb/dns -Rrsx /net.alt -f /lib/ndb/external #external - or - ndb/dns -s #internal aux/listen -q -t /rc/bin/service.auth -d /rc/bin/service.ext /net.alt/tcp - erik
Re: [9fans] Plan 9 via QEMU
On Fri, Aug 21, 2009 at 6:28 PM, Akshat Kumaraku...@mail.nanosouffle.net wrote: This time with just flag authdebug and without going through secstore, since it gave me the problem highlighted above, plain auth from the QEMU host gives the following at Ken FS: il: allocating ... user akumar = 2 authenticated authorize: uid is 2 authorize: uid is 2 hangup! connection timed out-3 ... i also see it trying to auth bootes (i think). do you have a user bootes? - erik
Re: [9fans] Issues with 2 networks, fs server, and namespaces
2009/8/21 erik quanstrom quans...@coraid.com: On Fri Aug 21 19:55:55 EDT 2009, devon.od...@gmail.com wrote: Well, we're getting somewhere. Using /cfg/cpu/namespace still seems to do nothing to get ether1 into /net. Putting it into cpurc does the trick though, go figure. you need it in both places, as namespace doesn't apply to the console. Aha. Ok. It's working now. Turned out the fs had its authdom in nvram set to the wrong domain. Oops. Thanks all! --dho - erik
Re: [9fans] Recursive structural expressions?
On Thu, Aug 20, 2009 at 9:14 PM, Charles Forsythfors...@terzarima.net wrote: qed allowed naming of regular expressions using `e' and their recursive invocation using \E, with results suggested earlier. http://cm.bell-labs.com/cm/cs/who/dmr/qedman.html http://cm.bell-labs.com/cm/cs/who/dmr/qedman.pdf ``It should be noted that the ability to define regular expressions recursively makes the term regular expression a misnomer: it is not hard to see that expressions can be constructed to match exactly the members of any given context-free language.'' I guess I missed this when I last read that paper. Do you know how qed dealt with infinite recursion or ambiguous CF expressions?
Re: [9fans] Plan 9 via QEMU
i also see it trying to auth bootes (i think). do you have a user bootes? Yes, bootes is the hostowner of the CPU/Auth server. I've got a QEMU image that was distributed by Devon, so I could at least try to get into Plan 9 inside QEMU and try to connect to Ken FS server from there. I'm still PXE booting, this time with a 9pcfil kernel that supports IL protocol and booting locally. It currently is just hung at time... fossil(#S/sdC0/fossil)... If I start QEMU with the option to boot directly from the HD image, as opposed to booting from network, then it starts up fine - but then the kernel is different also. I don't know what part of this is really troublesome. Maybe the pcap device? I can't get into VNC if I start QEMU with -boot d and -net pcap,devicename=ath0 so -boot n remains my only options; but with this option we have the hung fossil... or something So many problems (due to so many limitations). Cheers, ak
Re: [9fans] Plan 9 via QEMU
On Fri Aug 21 21:35:38 EDT 2009, aku...@mail.nanosouffle.net wrote: i also see it trying to auth bootes (i think). do you have a user bootes? Yes, bootes is the hostowner of the CPU/Auth server. also must have a user bootes on the fs even with auth disabled. might be good to turn on more debugging flags. like il. - erik
[9fans] ndb/dns as a slave
How do I designate ndb/dns to accept zone transfers from another one? I have dnsslave set to the other machine in the `master zone file' (for lack of a better term). The secondary server doesn't seem to accept updates. (Or maybe the master isn't pushing them? Dunno.) --dho
