mergemaster -U overwriting modified files
I recently began testing mergemaster -U since the perpetual review diff of file I never touched grows annoying real quick. Unfortunately I recently discovered that it does not seem to do what you might expect. For example it nuked my mailer.conf on one machine, and my /etc/namedb/named.conf (!!!) on another machine. Is this a bug or intended? What is the intended functionality of -U? -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller peter.schul...@infidyne.com' Key retrieval: Send an E-Mail to getpgp...@scode.org E-Mail: peter.schul...@infidyne.com Web: http://www.scode.org pgp3BoLI8OXVH.pgp Description: PGP signature
Re: Backing Up ZFS
I'm just curious at what others are currently doing to back up huge amounts of data. eg. 2TB and onwards. I'm using rdiff-backup and some scripting to backup ZFS snapshots. Other than the use of the ZFS snapshots there's nothing special about it. If your use case is suitable for rdiff-backup, using it with ZFS is a nice combination. For 2 TB+ I do suspect you would want to divide that up into multiple distinct rdiff-backup sets and indeed dependening on situation you may still have a problem with performance. In general I don't know that there is a lot ZFS specific to know about backups other than the availability of snapshots, and other than the potential to use zfs send/receive. Personally I am reluctant to use ZFS send/receive at this stage, because it is too dependent on ZFS. I would love to use it for maintaining a hot standby machine, or having an almost-realtime backup in the best case senario. But I would probably want a generic non-ZFS specific backup as my primary backup as well. One risk that you want to target with ZFS is that of a bug in ZFS itself; such bugs could conceivably be such that it affects your zfs send/receive backup. You mention: 3. ZFS - Remote ZFS using RSync (Living in Australia, there are limits on data transfer of a few hundred GB per month, to costs are prohibitive) rdiff-backup will be good in this senario too, giving you a rolling window of history in addition to an up-to-date mirror. It does do incremental updates including applying the rsync algorithm on individual files. It is definitely slower, in terms of CPU usage, than rsync however so if you have massive amounts of small files you may feel there is an issue. That said, I'm using it regularly to backup millions of files (e.g. collections of Maildir mailboxes). I mention rdiff-backup but of course there are plenty of others. I just happen to prefer rdiff-backup, mostly because of it's rsync mirror + history semantics and completely trivial setup. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller peter.schul...@infidyne.com' Key retrieval: Send an E-Mail to getpgp...@scode.org E-Mail: peter.schul...@infidyne.com Web: http://www.scode.org pgp3geOfKDD0f.pgp Description: PGP signature
Re: Copying system/ports configuration?
The last time i did this it took a lot of time to get the VM set up--installing all the right ports, getting the database configuration right, etc. Are there any shortcuts for this, e.g. a way i can automatically install the same ports on the new machine? I didnt see anything in the handbook or FAQ about this, but id think that people need to do this all the time. Any other advice for mirroring the system? Opinions definitely vary on what the best method is. What I do nowadays is to: * First of all, never ever interactively configure any ports. * Maintain any necessary global options in /etc/make.conf * Maintain port specific options in /usr/local/etc/ports.conf, and have it applied by ports-mgmt/portconf. * Keep a list of packages that I want installed (NOT including dependencies). I then build packages based on this information in a clean environment; typically a dedicated jail. Because all configuration is declarative, and nothing is a function of interactive dialogs, I know I can reproduce the package set + config at any time. The upgrade on the target machine is a matter of pkg_delete -a to remove all packages, and then re-installing all the built binaries. I have some tools to help me do this that are semi-public, but not really in a nice state at this time. Let me know off-list if you would be interested. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller peter.schul...@infidyne.com' Key retrieval: Send an E-Mail to getpgp...@scode.org E-Mail: peter.schul...@infidyne.com Web: http://www.scode.org pgpLyIgs9Hyg3.pgp Description: PGP signature
xorg gfx card: amd64+xvideo+decent 2d performance
Hey, I've been trying to find anything that fits the bill. My only requirements are: * That it works with amd64 (on FreeBSD, obviously). * That it supports Xvideo. * That it has decent 2D performance. * That it works with wide screen resolutions (so I guess basically modern hardware + modern driver). Unfortunately, the 'nv' driver has significant performance issues so nVidia cards are out. I love what ATI is doing with the radeonhd driver, but no Xvideo yet. I know Matrox used to be a good bet, but apparantly modern cards require binary blobs. I hear good things about Intel chipsets, but then I cannot find any cards that have them (only integrated on motherboards, which puts too much of a constraint on the choice of motherboard). Anyone got any recommendations? -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgpCf9QdpgTXd.pgp Description: PGP signature
Re: dump/restore don't work, handbook lies
dump -0af /mnt/d201gly-0.dump / [snip] restore -rf /mnt/restore/d201gly-0.dump it complains about '/' issues it complains about 'expecting YY got ZZ' I very rarely use dump/restore, but based on the man page I cannot see what's wrong other than the live fs issue already mentioned. Since no one has suggested the real problem, I would like to suggest that all those 'expecting ...' are also related to whatever errors were printed at the very beginning. So an actual dump of the exact output there would be useful. FWIW, for doing stuff like moving the root fs (which I have done more often than I would like) I recommend using tar -cp or rsync -a. I preserves everything I care about preserving, and it has well-known and well-tested semantics that I feel comfortable with. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgpR9EPx4KtGx.pgp Description: PGP signature
Re: ZFS-Pool is lost after reboot ( amd64 )
If I run zpool import x1 it works. But as you say it should do it by its own. Maybe it whould be the best to open a bugreport ? In addition to what has already been mentioned by Ivan Voras, make sure your /boot is not the subject to strangeness. Specifically, imported pools are kept track of in /boot/zfs/zpool.cache. If you somehow have multiple versions of your /boot, or it is read-only somehow, you may see problems like this. (For example, if your real /boot during boot is different from the location actually at /boot once the system has mounted all filesystems, and ZFS is loaded at boot due to loader variable.) -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: I'd like some help
I've been learning about a bunch of the BSD OSes, and i want to try Free BSD, but i can't figure out how to download it and the instructions don't make sense. I am running Windows XP OS- can you tell me how to download Free BSD with it? The easiest method is probably to download CD-ROM images as per instructions at: http://www.freebsd.org/where.html More specifically for amd64 (if you have one of the newer 64 bit computers) you have: ftp://ftp.freebsd.org/pub/FreeBSD/releases/amd64/ISO-IMAGES/7.0/ Or for i386 (older 32 bit): ftp://ftp.freebsd.org/pub/FreeBSD/releases/i386/ISO-IMAGES/7.0/ DIsc 1 (ftp://ftp.freebsd.org/pub/FreeBSD/releases/amd64/ISO-IMAGES/7.0/7.0-RELEASE-amd64-disc1.iso and ftp://ftp.freebsd.org/pub/FreeBSD/releases/i386/ISO-IMAGES/7.0/7.0-RELEASE-i386-disc1.iso respectively) are enough to install the operating system directly from CD. For actually burning the images onto a physical CD, you would have to use whichever CD burning software you have in Windows XP. Make sure you consult the FreeBSD handbook in relation to installation and aftewards: http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/ I don't know if this made any more sense than what you have already read. If things are unclear, please clarify what part you are having trouble will! -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: FreeBSD bind performance in FreeBSD 7
The people complaining about hardware compatibility need to pull their heads out. If they are buying brand new systems they are utter fools if they don't check out in advance what works and what doesen't. It's not like there's a shortage of experienced people on this list who could tell them what to buy. And if after the fact they find out their shiny new PC won't run FreeBSD - then they take it back to the retailer and exchange it for a different model. Why is this so difficult? The difficulty is not in checking out hardware before hand, the problem is FINDING hardware that satisfies your requirements. Just because I know that NIC so-and-so is recommended, it does not mean that I can find a complete server that: * Is within the budget. * Whose NIC is recommended for use in FreeBSD. * Whose disk/raid controller is recommended for use in FreeBSD - Including proper handling of write caching, cache flushing, etc * Is being sold in a fashion that is acceptable with respect to hardware support / replacement parts. * Otherwise is known to work with well FreeBSD. If you are a large company buying 200 servers I'm sure it's not a problem to get sample servers to try things on, or go for more expensive options just because of perceived FreeBSD compatibility. If you're a poor sod trying to get *one* machine for personal or small-company use and you want something that works and is stable, especially if you want it rack mountable, it is NOT necessarily trivial. Part of it is the problem of finding a solution that meets the requirements, and parts of it is about figuring out whether a particular solution DOES meet the requirements. For example, once your cheaper Dell server has arrived and you suddenly notice that it's delivered without a BBU, and clearly has write caching turned on based on performance, try asking (remember, this is a lonely customer with a single service) Dell hardware support whether that particular controller will honor cache flush requests right down to the constituent drives... I did, and eventually got a response after 1-2 weeks. But the response was such that I could not feel confident that the question was accurately forwarded to the right individual. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: How backup huge pgsql ?
I want to known how can I make backup of huge postgresql database (huge mean ~ 2To). I can stop the access of the database during N1 hours. Any idea about this ? pg_dump should work as usual. No need to stop database access since read-only access won't block anything else and it won't be blocked by other processes. That said, pg_dump:ing a database that is 2 terrabytes in size is going to take some significant time. If you want to maintain frequent backups you may want to look into using point in time recovery and WAL archiving. See: http://www.postgresql.org/docs/8.2/static/continuous-archiving.html Also the very long transaction used for the backup will prevent vacuuming from freeing tuples for the duration of the backup. If you have tables that rely on very frequent vacuuming for performance, those may be affected. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Howcome mail deletion time varies?
I use, for my imap-based mail, a combination of postfix, dovecot, thunderbird, enigmail (for gnupg), and openssl for browser security. When I delete mail messages, the majority of them delete (what seems to me to be) instantaneously, but a small minority of mails takes quite a bit longer, about maybe 20 seconds. Any idea what might be occurring on those mails, to trigger this really long delete time? Suggestion: Thunderbird does local compacation of the mailstor, and deletions that hit the threshold and trigger this may be slower than others. (I have a vague sense thunderbird does this; I could be wrong. Assuming the server is using Maildir I don't see why this should happen on the serverside, nor have I had that experience even with large folders (large enough that no mail client can handle them properly)). -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: dovecot questions
configuration isn't quite so simple. There were several in this list that suggested the use of Dovecot when I inquired for suggestions as to a good POP/IMAP server. I'll add my recommendation to that (having used courier in the past). So, I'm hoping that these same folks can offer help in getting me running. The WIKI for Dovecot is helpful, but seems to assume much in terms of what the reader will understand. How do I add users to the system? If I've done my work correctly, I've setup my Dovecot system to store virtual users in a PostgreSQL database. However, how do I add users? Dovecot doesn't really care. You provide the method of obtaining the user list from the database, in the form of SQL statements (assuming they differ from defaults), and it's up to you to make sure this returns the appropriate information (so in other words, dovecot doesn't add users for you). I used to run a pg+dovecot+postfix, but have since moved away from it and I don't have the configs easily accessible to check out specifics. But googling, this should be useful in terms of providing a bunch of real-world configuration examples: http://www.gjdv.at/snippets/linux/virtual_mail_hosting Also, with respect to the configuration file (dovecot.conf) why are there so many passdb/userdb? If I have passdb sql turned on, should I turn off all other passdb sections? What is the significance of the userdb static { } section in that file? It appears that it is necessary for use if using a single user to access several mailboxes (i.e. virtual users which is what I want to implement). Is this true? I can't answer each one of the above off hand, but things like uids is controlled by the user database that it sounds like you want to keep in PostgreSQL. This includes the ability to set the uid/gid, which you can have different for each user, the same, or some combination thereof, on a per-user basis. Your exact table design is up to you, as long as you can give dovecot the appropriate SQL statements for obtaining relevant information. If I'm understanding things correctly, the next, very important item, is how do I setup new users and how would those users then manage things like passwords, etc.? This is up to you. dovecot does not provide and user interfaces for managing accounts (that I am aware of). Typically a reason to have the user database in a relational database would be to enable the construction of such interfaces, or perhaps use of existing tools. But unless I am missing something, user management is beyond the scope of what dovecot itself is providing. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Updating ports
I don't understand this statement. I have killed portupgrade on numerous occasions, both locally and remotely, and have never had a problem restarting later. If you mean portupgrade doesn't restart where it left off, then yes, that's true, but only in the sense that it goes through all the ports checking for upgrades before returning to the build you left off at. Actually I was wrong because portupgrade doesn't do what I want at all to begin with, so because nothing was ever started correctly, there is nothing to resume correctly. The intended situation was: Mini-port tree contains: A B C D depends on C Now, C is updated in the tree. You issue: portupgrade -r C If all goes well, C is rebuilt followed by D. But if interrupted after C, D won't get upgraded on a subsequent run because portupgrade does not know C was upgraded. Of course, this is based on portupgrading doing that to begin with and AFAIK this is not the case. I am not sure if any such logic is possible at all in fact. portupgrade -rf C works of course, ***IF*** you know that C was upgraded. What I lack is a portupgrade -a --force-if-dep-was-upgraded, and even if that existed, the re-start problem would remain unless the fundamental approach was changed. (I have been meaning to fix this with my own package manager, but the project has been stalled for a while.) I *really* don't understand this. I can count on one hand the number of times that I've run into dependency problems with portupgrade, and all of those were addressed in /usr/port/UPDATING or by simply deinstalling and reinstalling the port in question. I would love to hear what I am doing wrong. I have just never ever had good experience with it. Everywhere you read on mailinglists or wherever, you have people recommending various versions (portupgrade -a, portupgrade -arR, etc) but none of it ever works over time for me. Firstly,there are these stale dependencies that are never explained anywhere as far as I can tell. I am also suspicious of the methology used that causes any kind of database / dependency inconsistencies as a matter of expected procedure. The job of the tool is to get my installed packages in synch with the ports tree; there is no possibilities for stale dependencies here as far as I can tell, except in some very specific cases. But everywhere I look in online resources these stale dependencies seem to be treated like some kind of unexplained-yet-necessary fact of life that nobody understands but that everyone seem to have a vague sense about. (I do realize upgrading is difficult in several fundamental ways; I wrote pkgmanager to do in-place upgrading for pkgsrc in a manner similar to portmanager - so I do have some experience with this. The re-write of pkgmanager to also support ports is what I refer to above. But with all the kinks of pkgmanager, the fundamental approach worked very well in practice, modulo some issues that have to do with lack of implementing particular cases, or fundamental problems in the underlying package management system.) Secondly there are various magic failures that start happening as a result of some dependency X being upgraded (or NOT upgraded) such that the other package Y depending on X breaks. This typically gets resolved by concluding that ok, it's all borked, I'll portupgrade -rf Y (or portupgrade -Rf X, depending)). Generally, these failures can be characterized as being such that they do not occurr if you 'make install' on a clean system with a consistent ports tree, but only occurr as a side-effect of problems with the upgrading procedure directly, or indirectly because packages are tested on fresh trees and do not stress dependency edge cases. Note that all this is specific to wanting to synch ALL packages. I never go around sniping at particular packages since i consider that to be a fundamentally broken approach in most situations. I just want to have everything upgraded to their latest versions (with security fixes), not have to micro-manage individual packages. Also, I do pay attention to /usr/ports/UPGRADING, but issues accounted for there definitely to not cover all the problems. Actually. Is there anyone heavly involved with ports that might be interested in discussing some of the issues having to do with upgrading? I have my own private little vision of what I want to see from ports/pkgsrc itself to enable package managers to support seamless upgrading. If there could be some cooperation going in terms fo enabling upgrading tools to work better, I might be more motivated to finally resume work on that pkgmanager rewrite. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Updating ports
What's the difference between them? The main difference that is relevant to me personally is that portmanager makes no attempt to be too smart about avoiding compilation, and it is fully restartable without affecting the results. It rebuilds ports in such a way that the result is, in theory, supposed to be equivalent to what you would have gotten had you installed them all from scratch with your current ports tree. In particular, given a re-build (e.g. upgraded) port X, all ports depending on X will also be re-built regardless of whether that is required according to the dependency relation. This is handled in such a way that it is not dependent on the entire procedure completing in one session, as you are with portupgrade (meaning it's restartable, as mentioned above). In practice, I find this is the most useful upgrading method. I have never been able to use portupgrade for more than a week or two on a real machine without running into issues (stale dependencies, failed builds due to weak dependency information, etc). That said, it's not perfect. The implementation is buggy in some ways, and there are fundamental problems with that upgrading approach (e.g., files moving between packages can cause problems). In the end I tend to either build binary packages from scratch and use portupgrade -afPP to upgrade, or do in-place upgrading with portmanager. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Apparently, csh programming is considered harmful.
other BSDs for that matter. It being GPL guarantees that quite apart from it general suckiness. Can someone please explain why bash sucks? Everyone keep's saying this but I have never heard anyone explain why, other than the GPL issue. I really want to know. (This is not because I'm a bash fan. My personal favorite happens to be zsh.) I tried replacing /bin/bash with /bin/ksh on a Linux system and it almost completely broke it. Suggests the Linux folks can't write boot scripts without bashisms. If this is a poke at the use of #!/bin/sh when the script actually requires bash, I 100% agree. However, if your intent (and the intent of Chuck Robey in that earlier) post is to imply that it's bad programming practice to write anything than POSIX compatible scripts, then I have to ask again - why? This is kind of a pet peeve of mine, so here goes somewhat of a rant. Please enlighten me as to why I am wrong: I don't understand why everyone insists on POSIX compliance for portability with shell scripting. The POSIX common demoniator seems to suck. Seriously. One keeps seeing things like: if [ x$var = xvalue ] When the intent is: if [ $var = value ] Because there is presumably some wonky script out there that breaks on the former (or perhaps its POSIX, dunno). I have recently began to appreciate that all this madness that would normally be considered unforgivable code obfuscation in anything but shell scripting, is all an attempt to somehow be portable. In any number of situations I would consider it much preferable to juse choose one particular shell and stick to it, rather than having to do battle with all these minor incompatibilities. Many major shells are very portable to begin with, and in many situation you *REALLY* don't care about some exotic Unix platform that 10 people in the world run, but where bash/zsh/whatever doesn't. Another example of the madness is: http://www.netbsd.org/docs/pkgsrc/makefile.html Check out section 12.3.3. Can anyone claim that it is sensible for it to be this fricking difficult *to print the value of a variable*? Although that last bit has to do with more than the choice of a shell, it highlights perfectly the type of trouble you run into when you try to be portable with the least common denominator. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Jail question
I want to check the understanding of jails. My understanding is a jail uses the existing kernel configuration and cannot use its own kernel configuration. Is this correct? Yes. The jail is being executed by the same kernel as the host system. The jail just has restricted access to certain system calls, which creates the sandbox. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: csh programing book
Well, I was only giving my personal opinion. I've never used irb, but it seems to me that using any sort of OO tool as a shell would be cruel and unusual, but I guess it takes all kinds, and I certainly wouldn't prevent you from enjoying yourself, same as I'd expect from you to mine. The OO nature of it really really does not come into play here as far as I can tell. Whatever you can imagine with perl/python/your-favorite-language is very likely as much possible, if not more so, with Ruby. Certainly if writing something elaborate it comes into play, but then we are away from the shell aspect of the discussion. Doing: system myfile does not become any more obnoxious just because myfile happens to be an object in a well-defined class-based object system. Nor does the function definition: def myfun(param) ... end Become obnoxious just because Ruby happens to be OO, for some definition of OO. My main concern is syntax. I would love to have a shell based on Ruby, Lisp, or some other powerful language (anything that at least allows functions to return values other than status codes... please). But I have yet to find one that makes it maximally simple and efficient to do the common stuff that you use interactively - which is to run external processes. scsh (Scheme Shell) comes pretty close; I have no objection to using it for shell *scripting*. Neither might Ruby be an issue with sufficient API support. But I am not sure about using it interactively. When interactive, you don't want to type even a single annoying character more than you have to. Or at least I don't. That said, scsh might be possible to tweak sufficiently. It actually manages to combine the power of Lisp with the convenience of shell scripting pretty well (as always, by using macros). So you have pretty low-overhead syntax like. For example, instead of: more myfile You have: (run (more myfile)) If you imagine an interactive mode where a top-level (run (...)) would be implied (under certain circumstances), you could make that be exactly equivalent to the normal shell version: more myfile This is true even with parameters; the macros are such that you need not explicitly make them strings (so (run (ls -l /)) is valid for example). I especially like the integration as soon as you want to do something slightly intelligent. E.g.: (run/strings (mystuff --list-something /path/to/db)) Yields an actual list of strings (one per line) that you can touch, pet and otherwise have your way with even if you want to do something other than piping it to the next process. Not to mention having higher order functions at your fingertips... In short, I would just love to have a single language for both tasks, not having to switch from one to the other after some threshold of script complexity. Unfortunately scsh has some issues (e.g., freebsd port is marked as broken om amd64 right now), so I dunno about counting on it being available everywhere. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Freebsd filesystem ( hard reboot )
Exactly, which is why I thought that just bypassing all those interventions with -y was 'brushing under the carpet'. No? Ah I see. Yes. Given that all bets are off, it's hoping for the best ;) I realise it would normally be excessively cautious to go for synchronous mounting, but what about for environments where power supply is such a major problem? If write caching is disabled (and confirmed to truly be disabled), it should not be needed. So as an added step beyond disabling write caching, it doesn't feel particularly useful. If write caching is still enabled, synchronous writes won't help except perhaps to lower the statistical probability of running into problems (that's just a guess). -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Freebsd filesystem ( hard reboot )
If you are running without write caching turned on (which is the default), That should be, if you are running WITH write caching turned on. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Freebsd filesystem ( hard reboot )
My understanding from the reading I have done is that in a situation like this where power outages are a danger (and presuably having the UPS signal the server to shut down gracefully is not practical), you need to make the file system as robust as possible in the first place, rather than rely on fsck -y after the event. Doesn't fsck -y rather sweep potential problems under the carpet? fsck is not sweeping potential problems under the carpet, as long as nothing unexpected goes wrong (software bug, hardware problem). The reason fsck works to begin with, is that it is designed to fix specific inconsistencies in the file system that are expected. The file system (takling about UFS here, and other non-journaled file systems that care about this stuff) is designed very carefully such that certain correctable inconsistencies happen, while preventing those that are not correctable. That is, under fully expected circumstances, UFS is intended to require fsck on reboot. But it is NOT intended that fsck find unexpected inconcistencies and ask for operator intervention. What happens in the event of write caching + power failure, software bug or hardware bugs, is that you end up with semi-random inconsistencies. fsck *may* be able to patch the situation enough for the file system to be usable, but fundamentally all bets are off. First step surely is to *disable* write caching if you have drives that are doing it? For UFS/reiserfs/xfs/jfs/ext3fs/ext2fs, yes. Then consider mounting the file system synchronously. Mind you, I don't know what the scale of the performance loss would be, and whether anyone does this nowadays! Synchronous mounting is not required for consistency (except perhaps for ext2fs; not sure). It is enough that the system does not break the file system's ability to guarantee ordering of certain critical operations, which is why write caching causes a problem (the drive re-orders writes for performance and you end up with B happening before A, but consistency depended on B happening AFTER A). -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: Freebsd filesystem ( hard reboot )
Well any number of things, but the most recent was a prolonged power outage. It is important to differentiate between expected fsck activity and unexpected. If you are running without write caching turned on (which is the default), a power outtage will constitute a crash from which a file system cannot guarantee to recover unless it takes measure to punch through the write cache at appropriate moments. If you machine shutdown softly as a result of the UPS communicating that power was running out, you should not have to fsck (barring other issues in the past). fsck need in these cases would indicate a software bug, or a hardware problem. If on the other hand your machine just lost power when the UPS finally died, you are relying on luck for recovery if you're on ufs/reiserfs/xfs/etc. Some environments will correctly handle this (e.g., ZFS), but most won't. The problem being that drive write caching will prevent the file system from guaranteeing ordering of certain critical operations that must be ordered in order to guarantee successfull recovery to a consistent state. That said, you are not supposed to need to answer interactive questions on boot for all cases of expected inconsistencies. If you are getting prompts as a result of unexpected inconsistencies, that indicates *something* is wrong. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: sysutils/etcmerge vs mergemaster
Isn't that what the -U option in mergemaster does? -U Attempt to auto upgrade files that have not been user modified. I believe I tried this once and it did not make a difference, but I didn't investigate. Perhaps I screwed up. But even so, three-way merging is nice, so etcmerge remains interesting. Thanks, -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
sysutils/etcmerge vs mergemaster
Hello, etcmerge, with three-way merging, has been available for a while, but the man page still warns of it not being extensively tested, and of course mergermaster still seems to be the officially supported tool. In spite of this, etcmerge is attractive since, to be honest, manually saying yes update to a bunch of files that mostly have only CVS revision changes is a waste of time. (In addition even files with local changes would be easier to handle with etcmerge) Given the obvious benefit to etcmerge's fundamental algorithm I have to wonder why it does not seem to be more wide spread in use. Are there problems with it that I don't know about? How many people use it in production? Are there advantages to mergemaster that causes mergemaster to even be preferred over etcmerge as the default tool in base? -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: This is a digitally signed message part.
Re: cp --verify?
cd /usr/ports/net/rsync make install clean rehash man rsync rsync -cav sourcefiles destfiles http://en.wikipedia.org/wiki/Rsync rsync is one of the great programs to discover. Note that one will have to run it twice; the first time to copy and the second time to verify that rsync does not want to update any files there (if it did it would be an indication of corruption). -c makes it checksum regardless of whether the file size/ctime matches, but it will not make it re-do the checksumming after an update. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgpHF8ZZI7HBx.pgp Description: PGP signature
Forcing buffer cache to use available memory
Hello, Is there a way to force the buffer cache to be more aggressive when caching reads? Or even just plain force a certain number of megabytes to be dedicated to the buffer cache? -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgpB9FEhefHvV.pgp Description: PGP signature
Re: Forcing buffer cache to use available memory
You want to adjust the vfs.read_max sysctl, I believe, or the vfs.maxbufspace for your second question. sysctl -d vfs is likely to be informative Thanks! That looks like what I'm after. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgp7jXUOdGw4S.pgp Description: PGP signature
Re: increasing maximum connections in PGSQL
Can anyone point me to some How-to documentation on raising this value to increase postgresql connections? http://www.postgresql.org/docs/8.2/static/kernel-resources.html -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgp0QxLEa4M4W.pgp Description: PGP signature
Re: Unable to delete a package
=== Installing for libgda3-1.9.102_3 === libgda3-1.9.102_3 conflicts with installed package(s): libgda2-1.2.4_1,1 They install files into the same place. Please remove them first with pkg_delete(1). As people have noted you can use -f to force deletion. The problem you are experiencing would typically happen either because you are trying to install some top-level packages that are truly in conflict, in the sense that they depend on conflicting versions of libgda. Alternatively, you have older versions of various packages installed that depend on the older version of libgda3, and are now trying to install the gnome stuff from a newer ports tree. In this case you may want to perform a full upgrade (using portupgrade/portmanagaer/portmaster/etc). forcibly deleting the package will work, but the other packages depending on them will very likely break. (As an aside, is it okay to run these installations from a terminal window from inside the Gnome GUI environment, or should I be outside of Gnome altogether when I do program installations/upgrades?) It doesn't matter, except in so far as any actions would kill your running terminal. I would suspect that the already running terminal would survive (certainly this is the case with most), though I don't know if the gnome terminal is doing stuff even after initial start-up that might cause it to die. (To be strict of course it's always safest to not run stuff you are actively upgrading, while upgrading. In practice though it tends to work.) -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgpqO0EpOXXKW.pgp Description: PGP signature
Re: Unable to delete a package
I'm using the command 'portsnap fetch update' each time before I do a 'make install clean' hoping that will cover me. I used portupgrade with 5.4 but switched to portsnap with 6.2 because I believed from the Handbook that it was a 'new and improved' way of maintaining my ports tree. Is this correct? Does portsnap do a better job than portupgrade or portsmanager? Do they all do the same thing (...in real life as well as on paper...)? Does portsnap automatically upgrade my programs or just the ports tree? And if I am using portsnap, can I use portupgrade or portsmanager as well, or will they cause a conflict? portsnap is the new alternative to cvsup for keeping your ports tree up to date. It will not keep your installed packages up to date however. Typically one would use portsnap for the ports tree, and then some other tool for package upgrades. The most official and traditional tools seems to be portupgrade (ports-mgmt/portupgrade). If /usr/ports/UPDATING has special instructions they tend to be for portupgrade. I have personally found portmanager to work better (ports-mgmt/portmanager), and have recently began trying out portmaster (ports-mgmt/portmaster). The redeeming feature of portmanager is that it attempts to recreate your package installation in such a state as you would have gotten had you done a clean 'make install' on a fresh ports tree with no packages installed, while portupgrade and, as far as I can tell also portmaster, tries to be smart and update only packages that have actually changed, and honor dependency information. I have found that in practice trying to be smart just leads to trouble (someone feel free to flame me). You may want to try both approaches. I prefer to do upgrades and installations from outside the GUI. But I've added 'gdm_enable=Yes' to my rc.conf file so that the graphical login screen appears for most daily use. In order to get to the pre-GUI terminal as root, I have to log in as root into the GUI, comment out 'gdm_enable=Yes' in my rc.conf file, then reboot the machine so that I can run portsnap (or whatever) in the pre-GUI terminal. Then after running portupgrade, I have to re-edit my rc.conf in order to re-enable gdm and reboot again. Is there an easier way to get the pre-GUI terminal without having to reboot after commenting out 'gdm_enable=Yes' in my rc.conf file and then re-inserting it after I do an upgrade? I've tried Alt-Ctl-Backspace, and that does kill Gnome, but then it just bounces me back into the graphical login screen. Ctrl-Alt-F1, f2, f3 etc will switch to the respective virtual console while X is still running. You can then switch back to X (probably at virtual console 7, 8 or something). If you want gdm to die you have to actually kill gdm; otherwise it will restart the X server when you kill the previous instance (thinking you just logged out of your session). -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org pgpHuh90UoABk.pgp Description: PGP signature
Forcing GEOM to re-taste a device
I am having trouble with a USB stick (a Verbatim store'n'go, 4 GB). It seems there is a timing problem. On insertion there are complaints that there is no medium present when attempting to discover the size of the device. It goes on to retry but eventually bails out. However sometimes (only once so far) it manages to successfully retry and discover slices; presumably due to timing. So I would like to force GEOM to re-taste the media (camcontrol rescand da0 is not enough). Is there a way to do this? And further, I was hoping to boot off of this. If anyone have suggestions as to how to make the retries continue for a longer period (other than patching the source), it would be welcome, since during boot I need the kernel to be able to taste it on the initial attempt, since failure will cause a panic immediately. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Forcing GEOM to re-taste a device
cat /dev/null /dev/da0 That should retaste the device. Thanks! For the archives, I also discovered that the issue can be worked around by physically timing the insertion of the stick. If you insert it just enough that it gets powered up (presumably initializing) but not enough for it to attach to the USB bus, wait a few seconds, and insert it all the way - it will get properly detected. Similarly if connected on boot there is no problem. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Dell 2950: 4GB not seen (amd64; works on other 2950:s)
You need to look closely at the hardware configuration for these servers and their motherboards. Often some memory is reserved for things like onboard video, etc. You can free up that video memory by adding a separate video card, but necessarily other memory that may be used by the motherboard. Unfortunately with dell systems same model's don't necessarily mean same motherboard. Also, how memory is used via the BIOS is dependent on the BIOS version. You should try to be sure all systems you want to compare have the same motherboard and chipset and that these also have the same BIOS version. I'll have a closer look and inquire with Dell what the intended functionality is (hopefully it doesn't turn out to be something that requires Windows/Linux to work around). Thanks! -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Dell 2950: 4GB not seen (amd64; works on other 2950:s)
Derek triggered a thought ... I believe the 2950s have the ability to do RAM RAID1, to increase RAM reliability. If that belief is correct, it could be that you've got 4G physically in the machine, but only 2G logically available to the OS. At least, I think I remember seeing an option like that in a BIOS ... Now that you mention it I do think I recognize that. It would fit with the ~ 2 GB visible memory, but on the other hand the kernel does print the full 4 GB during boot. But will definitely have to look into that. Thanks! -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Dell 2950: 4GB not seen (amd64; works on other 2950:s)
Hello, I have a Dell 2950 where my 7-CURRENT amd64 FreeBSD does not see all visible memory. It has 4 GB of physical RAM. dmesg on boot includes: usable memory = 4280811520 (4082 MB) avail memory = 4117716992 (3926 MB) Yet summing memories visible in top yields ~ 2100 MB. Of note is that I have 6.2/amd64 on several other 2950:s with 4GB of RAM which say on boot: real memory = 5100273664 (4864 MB) avail memory = 4122443776 (3931 MB) But has all 4 GB visible in top. Unfortunately I failed to notice this until after the machine has begun being used, so I have limited possibilities for rebooting/mucking with BIOS settings. I was hoping someone could suggest something right off the bat. In addition on the problem machine the following sysctl values are present: hw.physmem: 4280811520 hw.usermem: 3628220416 hw.realmem: 5100273664 hw.cbb.start_memory: 2281701376 hw.pci.host_mem_start: 2147483648 With hw.usermem being slightly higher (but not 2 GB higher) on the 6.2 system without a problem: hw.physmem: 4283285504 hw.usermem: 3998797824 hw.realmem: 5100273664 hw.cbb.start_memory: 2281701376 hw.pci.host_mem_start: 2147483648 I was under the impression that memory visibility issues were a thing of the past on amd64. Any insight? Thanks! -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Custom builds from ports
Is there a way to specify which ports certain options are to be applied to, without having to craft custom command lines and build ports individually? Check out ports-mgmt/portconf, which allows you to set per-port options in a manner independent of which tool you use for installation and upgrades. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD-6.1 bootup hangs after power failure
One thing I find FreeBSD very fussy and sensitive in comparison to Linux OSes is that whenever there is an power outage, something wrong is bound to happen. Maybe, it was made to happen this way but living in here over the other side of the world, we do have to face power outages despite our best efforts. The problem with power outtages is that in the presence of write caching *all* bets are off. In practice I have found this a bigger problem with UFS2 than with e.g. reiserfs on Linux; possibly because the characteristics of disk i/o in those cases make it less probably to actually trigger a problem in practice (but this is just my speculation, but fits my experiences). But regardless, all bets *are* off in the event of a power outtage. Unless you have battery backed caching controllers, you will need to disable write caching (hw.ata.wc=0) in order to be safe - at the cost of performance. Alternatively for 7.0+ or CURRENT you can use ZFS which understands these things and will actually send cache flush commands to the drives on transaction commits, thus allowing safe operation in the event of a power failure, while not taking the performance hit associated with disabling write caching. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Starting again from Scratch
4) Find out how to keep updated / informed on updates to packages I have installed, and do so where necessary. If anyone can advise, point me in the direction of tutorials or step by steps, on the above It would be greatly appreciated. I would like to understand why its not so great to just install everything from pkg_add, whats the advantages of ports etc Sorry, I don't know of a HOWTO specifically addressing it all like that. But I can say that I always use pkg_add -v -r on initial installs (for speed), and *then* upgrade using portmanager. Lately I have started using a jail for building binary packages of everything I want installed, and then doing a global upgrade by removing all packages and installing the binary packages built in the jail. It's fiddly, but works well in the end, and avoids problems you can run into with portmanager as well as minimizing the time during which your machine is not fully populated with packages. As for portupgrade, I have honestly never understood how anyone manages to use it for upgrades without difficulties. Whenever I try I run into problems almost immediately, having to do with packages not getting rebuilt even though they should and/or stale dependencies and whatnot in the pkgtools package database. If someone has magic information here I'd love to hear it. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
mfi RAID monitoring
Hello, the mfi(4) manpage briefly mentions the /dev/mfi? device, and that it provides a simple management interface. Is there any way to monitor the health of an array (the only thing of interest in this case) using this interface, other than getting the proprietary Linux blob to work? (I could not get the latter to work.) -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Raid Controller
I am building a file server off freeBSD, and am wondering if my sata raid controller is supported by the OS. I have the Addonics *4 Port RAID 5 / JBOD SATA II PCI Controller (ADSA4R5) controller host. The chipset is Silicon Image Sil 3124. If this raid card is not supported, can I install drivers once the system is running? Where would I find them for freeBSD? The 3124 is listed as supported in the ad manpage. However, in general a lot of people (including me) have had troubles with Silicon Image chipsets (google(freebsd sil3112 timeout) etc). I have had trouble with at least the 3112 and the 3114. I recommend some Googling on the specific situation with the 3124. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Stable many-port SATA controller recommendations
Still looking encouraging with 7 drives plugged in. I have had a couple of issues that may or may not have anything to do with the controller support. I will send an update in a few days or so when I know more. So it's been a while now and though I had a couple of weird pauses (console work, but any vfs access seemed to hang for several seconds) a couple of times in the very beginning, I don't know if they had anything to do with the controller. I have been running for almost a week now, doing rsyncs, rdiff-backup:s and generally mucking about fairly significantly. Aside from the above mentioned problem that happened way in the beginning, I have not had a single problem with it. Definitely my new favorite controller... -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: vidcontrol: getting active vty: Inappropriate ioctl for device
vidcontrol: getting active vty: Inappropriate ioctl for device I have this entry in the ~/.bash_profile file: vidcontrol lightcyan vidcontrol is trying to affect the system console, and does it by manipulation file descriptor 0 (stdin). If you are logging in at the user in question at the console it should work. But in X it won't. You can make it do what you ask to the system console with: vidcontrol /dev/console But that will require root privileges. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: In regard to SATA controllers.
SATA controller and continue using geom_raid3? This particular card uses a Marvell controller, so I would guess that it would be detected by FreeBSD without issue. If someone could confirm this, it would be much appreciated. See the stable many-port SATA controller thread. So far it's looking good for me, but it has not been that long yet. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Best remote backup method?
Also, dump/restore allows you to use snapshots on a live filesystem (I would test it properly on a large FS with heavy activity). But it's worth pointing out that this is fully possibly with any backup tool - just run mksnap_ffs and backup a mounted snapshot. I do this with rdiff-backup for example. Now, if you are worried about backing up the whole filesystem...well, just tell dump not to dump it :) man chflags (in particular, the nodump flag) man dump (in particular, -h ) The problem with this is for me two-fold: (1) It's a global property. I can't take different backups that include/exclude different things. (2) I can't easily express backup /usr/var/db/my-important-database without seting nodump on a bunch of stuff except that. In other words, I want exclude by default, while dump and the chflags system provides include by default. That said I do like dump's integration with snapshots and overall coherent feeling. If backup diskspace and bandwidth was not a concern I'd use it. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Best remote backup method?
I'm presently backing up two servers in a remote location to a usb drive located elsewhere by using rsync over ssh (all three are FreeBSD boxes.) After the recent discussion about dump, I'm wondering if I would gain anything by using dump rather than rsync. Has anyone used both? Any thoughts as to which is better and why? The rsync command I use is: rsync -avz ${LOCALDIR} -e ssh -i ${KEY} ${REMOTEHOST}:${REMOTEDIR} Personally I never find dump/restore practical since I seldom want to backup entire filesystems for performance/diskspace reasons. I have not found any truly perfect solution; what I am using most often is rdiff-backup which combines the properties of incremental backups and rsync. It keeps an up-to-date mirror along with reverse diffs. The good parts are that: * It works without hassle right off the bat over ssh. No fuss. * It is able to backup ownership information without running as root, because meta-data is stored separately from the files (but at the same time the up-to-date mirror is a plain tree on disk so you do not actually have to use rdiff-backup for restores unless you care about ownership and such). * Other than using the rsync algorithm for transfers, the actual reverse diffs are also expressed at a more granular level than entire files. End result is that a daily backup of that 5 gig log file will not consume 5 gigs of storage per day (but will be very slow to backup). The main downsides are IMO: * It's fairly slow. I don't generally see it saturating neither disk nor networking. Sometimes it's CPU bound, but oftentimes it's just slow without an obvious bottleneck (probably architectural in the protocol). * It has some reliability issues. A common problem is that certain meta-dat is kept in gzip files, and in certain cases of rdiff-backup being interrupted these files will get corrupted and rdiff-backup will refuse to function due to the gzip library throwing an exception. * While it basically works like rsync with history and is thus very simple to get started with, it does just that. If you want things like automatic rotation schemes with hourly/daily/etc you have to script that on top. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Stable many-port SATA controller recommendations
I'm definitely interested in hearing your results. Still looking encouraging with 7 drives plugged in. I have had a couple of issues that may or may not have anything to do with the controller support. I will send an update in a few days or so when I know more. Also, what is the model of the drives you're using with this card? 6 Seagate ST3320620AS (320 GB, SATA I) drives and one Maxtor 7V250F0 (250 GB, SATA II). Note that it is a PCI-X card. It works fine in the plain PCI slot on this motherboard, and the place I bought it from said they had tested it in at least one PCI motherboard. By your name you may be in Sweden; so FYI I purchased these cards from Mullet (www.mullet.se). They're the only supplier I have found that carries this locally. It's a shame really since it's such a good card given the price. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Stable many-port SATA controller recommendations
WARNING - WRITE_DMA48 UDMA ICRC error (retrying request) LBA=a number please post dmesg! http://distfiles.scode.org/mlref/freebsd-dmesg-promise-tx4-7current.tx Be aware that some disk drives do the wrong thing. Western Digital desktop disks, for example, don't simply remap a sector when they find a bad one, they sit there scrubbing forever trying to reread the data. WD calls this a feature and their server-quality WD disks that cost more don't do this. It makes use of these disks impossible in a raid array. Seagates and Maxtor desktop drives to my knowledge don't do this. This was Seagate. I have not investigated whether I can trigger it as easily with the Maxtor. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Stable many-port SATA controller recommendations
Are you seeing this same problem on both the Promise and the Silicon Image? No; with the Silicon Image I am getting the timeouts that everyone else is getting (google(freebsd sil3112 timeout). Have you tried a different brand of disk? Some combinations of controller and disk don't work correctly. With the Silicon Image I saw it with both Seagate and Maxtor. With the Promise I am not sure; might have been the Maxtor once or twice, but only the seagate when I triggered the bug on purpose (didn't try the Maxtor). If I use up a slot, I'd like at least 4 ports and NCQ. If FreeBSD doesn't have NCQ support yet I might just get a USB or FW to SATA adapter or two and wait for NCQ. NCQ is nice but I don't *really* care. I just want something that works at all :) Of course I would want NCQ if it was a performance critical production system though. And AFAIK NCQ is not yet supported in FreeBSD. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Stable many-port SATA controller recommendations
My main candidate is the AOC-SAT2-MV8. Can anyone offer input on the stability of this card in FreeBSD? It would be perfect because it is priced very well. I ended up getting an AOC-SAT2-MV8. Preliminary results are encouraging but I have not yet run with it for that long... I'll try to remember to post an update for interested parties and/or the archives when I have tested it more. -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Stable many-port SATA controller recommendations
Hello, after spending time and money on several Promise and Sillicon Image based cards that don't work properly, I am still in desperate need of a stable SATA controller. The goal is ONLY to (1) be stable and (2) get as many SATA ports as possible. I don't care about saturating the PCI bus or anything like that; this is for home use. I want stability and preferably cost-effectiveness. I don't care about RAID support since this is for use with ZFS. My main candidate is the AOC-SAT2-MV8. Can anyone offer input on the stability of this card in FreeBSD? It would be perfect because it is priced very well. If the Marvell is out, other main contenders are the 8 port LSI MegaRAID, Highpoint RocketRaid and 3ware cards. But these are all on the expensive side, with the half-exception of the MegaRaid 300-8XLP which is only semi-expensive. Any recommendations? -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: Stable many-port SATA controller recommendations
Promise are supposed to be one of the better makes when it comes to documentation and open source support, Yes, that was my impression prior to purchasing the two TX4:s. what problems are you seeing? I can easily (dd if=/dev/zero of=testfile bs=$((1024*1024)) count=500) trigger within a few seconds: WARNING - WRITE_DMA48 UDMA ICRC error (retrying request) LBA=a number Followed perhaps 30 seconds later by the usual flooding of g_vfs_done() errors and a panic resulting from the inability to flush dirty buffers. If anyone is interested I can provide actual details, but there's nothing really interesting to see as far as I can tell. I have tried several different SATA cables (since in other contexts with other controllers these messages were supposedly often indicative of bad cabling). The TX4, or at least this particular type, is specifically a PDC40718. I did talk to someone on ##freebsd who indicated he had not had any trouble with this particular chipset, so I presume it is somehow being triggered by the environment. It's an Epox EP-8KMM5I motherboard. It's supported, and even better, the Marvell is supposedly now 64bit clean with ata(4) in HEAD. When I tried it a year or so ago it would, like any other ATA controller, explode with more than a couple of disks and =4GB RAM because of problems with the way bounce buffers are allocated; I didn't see any problems aside from that, but I didn't test it very much. Ok. 64bit/4 GB is not an immediate problem for me, though it will be relevant in the future. But nice to hear someone had it working stably. When ata(4) proved itself utterly useless in environments with decent amounts of memory, I got an 8 port LSI MegaRAID, since amr(4) had just been given an overhaul to make it MPSAFE and 64bit clean; I've not had any real problems with it (aside the occasional fight with the BIOS, which is par for the course with RAID controllers), but I'd still prefer a real SATA controller so things are a little more predictable and flexible (SMART support, software RAID, disks exposed to the actual OS and not some dodgy half-assed ASIC..). Yes. Given that I do not intend to utilize the hardware RAID support, having to go for a RAID card is to me just a downside, even disregarding the increase in cost. Thank you for the input! -- / Peter Schuller PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org signature.asc Description: OpenPGP digital signature
Re: FreeBSD VPS providers
http://www.johncompanies.com/jc_bsd.html which seems like a good setup. For comparison you may want to have a look at leaseweb (www.leaseweb.com). They are offering dedicated physical machines for comparable prices. My private colo box is hosted there and so far my experience has been good, though I have not yet had any kind of hardware disaster or similar so I cannot speak to how their tech support works in those cases. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Empty directory 60M in size; used to contain 1.7 million files
Directories are only shrunk when a file is created and the slack directory space can be trivially truncated. This is to avoid useless compaction during rm -rf-style activities of a directory that will just be deleted anyway. Just create a dummy file with touch and the directory will shrink down to 1 block. You can then remove the file. Thanks, that explaints it. And it did work. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Empty directory 60M in size; used to contain 1.7 million files
Hello, Observe: hyperion# ls -la total 61634 drwxr-xr-x 2 xxx yyy 63047168 Nov 18 21:33 . drwxr-xr-x 6 xxx yyy 512 Oct 8 16:39 .. hyperion# find . . hyperion# The one special circumstance is that the directory previously contained 1.7 million small files, that are now deleted. This is on FreeBSD 6.1 with UFS2 + softupdates. No snapshots exist of the filesystem. 1.7 million files may be extreme, but I don't see why an empty directory would ever consume more than one inode? -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: pf + ipv6 + keep state - any known issues?
Are you using antispoofing rules on your external interface? If you've got something like this in your ruleset: antispoof log quick for $ext_if Then it will expand into a series of rules containing the following when you load them: Thank you for responding! No, this is not the issue. I *am* performing antispoof on my physical interface, but not on the tunnel interface. After some further investigation my current theory is that I have run into the trouble with pf and a packet traversing an interface twice. Having a 'keep state' on the *incoming* direction results in a state entry according to pfctl. But no state entry for the 'keep state' in the outgoing direction. The result being that while packets coming into port 22 are allowed and state set up, but the responding packets (to some random source port) are NOT allowed because the outgoing direction yielded no state entry. I am not sure what the behavior is supposed to be with a packet traversing the same interface twice, except I have seen references to the effect of don't be stupid, don't do that, get another NIC (for the typical firewall/gateway case). Except in this case that does not apply, even if you agree with the sentiment to begin with. Can anyone confirm or deny whether double traversal *IS* supposed to work without difficulties/special cases on current versions of pf/FreeBSD? Thanks! -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
pf + ipv6 + keep state - any known issues?
Hello, I am using pf on a 6.1 machine. I have a tunneling interface (gif0) for my IPv6 feed. The problem I am having is connecting to myself in spite of firewalling. I am allowing traffic on port 22 to my public ipv6 address. I am also allowing all outgoing traffic on the tunneling interface, with 'keep state'. ping6:ing myself works, but connecting to port 22 does not. The intial SYN gets through and is responded to by an ACK, but that ACK is seemingly dropped. This inspite of the fact that 'pfctl -s state' shows a tracked connection for the relevant port pair. I can work around it by allowing all packets from my own IP on the tunneling interface, but as far as I know this should not be required. That is, connection tracking should be working even for local connections on a particular interface - correct? Note that connecting to port 22 works perfectly from outside IP:s (I had someone external verify this) without any special casing of the rules. That is, I only have the usual rules for allowing the incoming packets to port 22, and the rule allowing outgoing packets with 'keep state'. The fact that this allows successful establishment to port 22 by an external party suggests to me that I have not made some trivial misstake in the rule - yet connections to myself do not work. My question is whether there are any known issues that this sounds like - or of course if there is some reason why this is not supposed to work by design. Thank you, -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
geom_eli and Safenet/hifn crypto accelerators
Hello, I am looking to purchase a supported crypto accelerator. Not for performance, but to off-load the CPU. The soekris hifn products are cheap and easy to obtain, but Googling yields a lot of posts that indicate there may be problems with the hardware. Safenet seems to be the other option (though I have yet to find a reseller) - but I have not found a single post anywhere about anybody's experience using it with geli in particular, nor FreeBSD in general. Does anyone have any recommendations regarding this? All I care about is low CPU usage and stability; performance is secondary. As a bonus, if anyone has CPU usage benchmarks for geli I'd be interested in seeing them. I did find one comparison while Googling, but it was completely inconsistent with what I am seeing. I am seeing complete CPU saturation doing sequential I/O on a USB device at about 10 meg/sec. (Actually 50% system utilization on a dual-core 3800+, which I take to be one core saturated unless geli is able to use both CPU:s for the same userland I/O operation.) Thanks! -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
pfsync broadcasts without explicitly enabling it
Alright, I have seen this on two machines now. At first I ignored it, but now I am wondering. Suddently I will spot an extraordinary amount of network traffic and on tcpdump:ing discover my machine is spewing out pfsync broadcasts. On both machines pfsync was NOT enabled, nor had it ever been. pf_enable=YES and pf_log=YES was in rc.conf, but nothing about pfsync. I can squish the broadcasts easily enough with 'ifconfig pfsync0 -syncdev down', but I am somewhat concerned over it being mysteriously enabled. One of these machines is a 6.1, the other a 6.0 (though it may have been a 5.x at the time it happened, not sure). pf is being used actively without any 'no-sync' options present; but it is my understanding that no pfsync traffic should be generated until pfsync is explicitly turned on with ifconfig. Does anyone have an explanation for this? I am quite sure I have never enabled pfsync in rc.conf, nor have I ever ran any ifconfig commands that might enable it. I am not running any kind of meta-firewall scripts or anything like that; there should be nothing but the standard FreeBSD rc scripts modifying network interfaces with ifconfig. The second machine I saw this on has only been running for a few days, so I am even more sure I haven't done so on this one than on the other one. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: background fsck
what is the rule to decide if filesystem may be background checked or not? for example my / is checked foreground, while /home checked background. can't root partition be background checked too? Last time I checked (5.3:ish), no. I have no clue whether it's intentional. I tried to fix it but never got too far (well, I gave up). If I remember correctly (which is not at all sure since this was a while ago and I did not make notes), it has something to do with the fact that the root filesystem is mounted read-only, and the appropriate flags that are required for background fsck aren't set until a dirty filesystem is mounted read/write. (But again, I may very well be mis-remembering or have misunderstood the code in the first place.) Ever since then I have started partitioning with a small root partition (a few hundre megs, up to a gig or so), a huge /usr and symlinks for /tmp and /var - as an alternative to just one huge root partition. Does anyone want to speak up as to whether the behavior *is* intentional, or just an unintentional bug? -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: gmirror, gvinum or ccd to mirror root-filesystem under 6.0R
i plan to install 6.0-R in near future and ask myself if i should use gmirror, ccd or gvinum (again) for software-raid for mirroring the root file-system, as to: - reliability, stability issues - performance issues - minimum installation/configuration effort - advantages / disadvantages of gmirror vs. ccd vs. gvinum what are the experiences here ? Personally I currently do not trust vinum at all (any and all of my edge case tests / simulated hardware failures have turned into disasters). ccd I haven't tried, but I have set up root-on-gmirror on at three machines so far. I am very happy with gmirror; I have only observed two major problems so far. Firstly, geom/geom_mirror seems to obtain an exclusive open of the drive. this makes it a royal pain to update the boot sector of a drive while the system is booted with geom having claimed the device (and it doesnt help that boot0cfg does not report the error properly (and the patch i sent has been ignored so far)) Secondly, on at least one occation, the total failure of a mirror (rebuild test and the drive being rebuilt FROM had a bad sector) resulted in a kernel panic. The filesystem was mounted at the time, so I presume this isn't a problem with geom_mirror per se, but rather has to do with an attempt to access a destroyed geom or similar. (This wasn't the root filesystem btw - if it was the root filesystem then the system has a right to panic :)) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Necessary code or trash?
Why most of bin (sbin) utilities are so big. For example, rm - 410 268 bytes, mv - 407 568 bytes, date - 423 748 bytes. Do they really contain only necessary code or have more than a half of trash? Presumable they are statically linked; I believe 5.0 falls before the switch to a dynamically linked /bin by default. (See eg http://kerneltrap.org/node/1628) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
3c905C: very high system/interrupt CPU usage
Hello, I have a machine here with a couple of 3c905C:s in it. For some reason, network activity causes very high amounts of system and interrupt CPU usage. By very high i mean more than 50% interrupt, and more than 25% system usage. It is not just a measuring issue either, because the whole reason I looked into it was that file transfers have been slow for as long as I can remember. Especially with ssh, everything is extremely slow given that only a small fraction of the available CPU resources are left over for userspace. Any ideas on what would cause this? I am not experiencing any other issues with the NIC or anything else, its just the high CPU usage. Could it be that the xl driver has fallen back to not using DMA (is that even supported with 3c905c?)? If this is the case, is there a way to find out whether it is currently using DMA? (I see nothing in particular in the kernel log). Also: The CPU usage problem occurrs on both outgoing and incoming traffic, and on both NIC:s. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 3c905C: very high system/interrupt CPU usage
I should add that this is currently on RELENG_6_0, but the situations has been the same for a long time. At least as far back as 5.3, but I believe forever (meaning probably 5.1 or so for this machine, I am not sure). -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: gbde - destroying master key without lockfile
Instead of destroy I use nuke. Thanks! -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
gbde - destroying master key without lockfile
Hello, I would like to use gbde to encrypt some disks. Using an external lockfile things work pretty much as documented (except for some options that aren't supported by the tool, but which are listed in the manpage). However, for this particular situation, I do not want to use an external lockfile. The manpage seems to imply that without -L/-l, the first sector is used as a lockfile. Indeed, I can init, attach and detach devices without an external lockfile. However, when I attempt to destroy the master key: # gbde destroy /dev/label/storage304 Enter passphrase: Opened with key 0 gbde: No -L option and no space in sector 0 for lockfile Trying to use -L for this particular operation fails: gbde: illegal option -- L Usage error: Invalid option And trying to specify -n -1 as the manpage says also fails: gbde: illegal option -- n Usage error: Invalid option So the question is - how do I destroy the master key (other than dd if=/dev/zero of=...) when not using an external lockfile? (The reason I do not want to use an external lockfile is simply that I do not see a need for it in my situation and I would feel much more comfortable if the gbde volume was self-contained; no need to backup anything else or keep it in synch.) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: gmirror rebuilds on every reboot
As per subject: I have a 5.3 box with two SATA drives and I using gmirror to achieve RAID 1. The problem is it started rebuilding the array *on every boot*. ... I cannot think of any reason this might happen, but I'd appreciate any hint to avoid that. First thing I'd check is the shutdown sequence. I dunno if this is machine you're physically in front of, but if you are, look for kernel output at the end of the shutdown sequence. The fact that it is rebuilding is usually indicative of an unclean shutdown. If it happens all the time it might be that there is some problem preventing buffers from being flushed properly and/or the gmirror to shutdown correctly. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: reload rc.conf without rebooting
How does one cause rc.conf to be reread without rebooting? Under linux I would generally do source somefile.conf. But if I do source rc.conf, I'm told that all my settings are not commands. Generally, you don't. For details, see the explanation of the FAQ: http://www.freebsd.org/doc/en_US.ISO8859-1/books/faq/admin.html#REREAD-RC It's worth mentioning that if you're on 5.x and are looking to reload certain specific settings (such as daemon configurations etc), you can run /etc/rc.d/affected-script reload (or the equivalent in /usr/local/etc/rc.d). -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Blocking on multiple threads with timeout
I have a few threads that might need as long as a minute or more to complete and terminate. If they exceed an arbitrary time, they can be canceled. In Win32, there is a 'wait on multiple objects' call. I'm not sure if it blocks or spins, but it *does* take a timeout argument. How about having the threads signal their completion explicitly while also signalling a condition. You could then have a monitoring thread using pthread_cond_timedwait() which would enable you to time out without busy-waiting. Cleanly killing the thread in question is another matter though... -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: glabel - refuses to label = g partitions
If I do: glabel label somelabel /dev/ad1s1g geom_label labels /dev/ad1 instead of /dev/ad1s1g[1]. However labeling /dev/ad1s1{a,b,d,e,f} worked fine. But /dev/ad1s1{g,h} does not (and probably not the rest above h either). Any idea what to do about it? The the purpose of the mailing list archive in case somebody else stumbles across it: I finally realized what the problem was. It had nothing to do with partitions with names g and above. The issue is that the information that geom_label looks for when tasting a device is stored in the last block (or close to it - at the end of the device, I don't remember). My last partition ended at exactly the last block of the device. Presumably /dev/ad1 is tasted by geom_label before /dev/ad1s1{g,h,whatever}, which means that geom_label will find the label on /dev/ad1... -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
glabel - refuses to label = g partitions
If I do: glabel label somelabel /dev/ad1s1g geom_label labels /dev/ad1 instead of /dev/ad1s1g[1]. However labeling /dev/ad1s1{a,b,d,e,f} worked fine. But /dev/ad1s1{g,h} does not (and probably not the rest above h either). Any idea what to do about it? I did some cursory checks to make sure that the glabel tool does not mangle the name of the device. The name does seem to propagate down unmangled all the way to g_metadata_clear() and g_metadata_store(). After that I'm not sure how the changes are picked up by the kernel, so I stopped. [1] I.e., glabel list reports what you would expect after a glabel somename /dev/ad1, and the kernel log contains: GEOM_LABEL: Label for provider ad1 is label/somelabel -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD 5.3 - Raid
I have read the manual and searched the web for a simple way to do the above. The manual seems to cover complex solutions and may be somewhat behind the times. Personally I would go for geom_mirror. See gmirror(8) ('man gmirror') for usage instructions including examples. Creating a mirror takes only one command. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Migrate from Qmail to Postfix..
I want a real confirmation that you can copy the Maildir folders and move them to a different box with postfix running with the ability to continue as it was provided that the postfix was configured to use Maildir and also the domains that this mail server should accept for? Postfix does support Maildir. I've used Maildir+Postfix+Maildrop aswell as just Maildir+Postfix where virtual mailboxes are in maildir format. IIRC its just a matter of having the trailing slash on mailbox paths. The biggest problem with a qmail-postfix transition is probably converting all the logic you may have encoded in all those .qmail files (if any). -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: POP3 IMAP
Which IMAP and POP3 ports are stable and good? Any reccomendations? I recommend dovecot. Like somebody else mentioned it's faster than courier-imap. And it supports POP3 and IMAP (only one daemon to configure rather than two). Currently administering a postfix+dovecot+postgresql setup that works well. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: portupgrade vs. portmanager
Right now portmanager is upgrading kdelibs and I'm still using it. The only problem that might occur is between the deinstall/reinstall steps I'll be missing the libraries for about a minute, when this happens I just wait untill its finished reinstalling then continue. Here is a recap of what happens: ... I've been testing this for a year now and haven't had a problem yet using a program while it is being updated. That does indeed sound perfectly painless. I failed to remember that there is another factor playing into my annoyance with pkg_chk - namely that it seems to remove dependent packages in chunks prior to re-installing them. So the end-result is that the system is completely missing several packages for extended periods of time while a bunch of other packages are being compiled. I don't know why this is done or exactly under what circumstances. But indeed, the way portmanager behaves as you described above sounds a lot better. Portupgrade has one serious flaw in my opinion and that is running something like pkgdb -F damages the port installation database as far as I'm concerned. It causes the data base to say ports were built with dependency ports that they were never really built with. Portmanager only addresses that one issue and for the forseeable future that is where all the focus will be, only on correctly updating ports. Ok. That sounds good. To be honest I have never understood why the dependencies seem to randombly break, requiring a lot of fiddling with pkgdb. I tried portmanager in the past but missed the fact of 'proper' upgrading of entire dependency chains. I will have another go now though. Thanks for the clarifications! -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: portupgrade vs. portmanager
That is indeed the case with portmanager. Sometimes it is a waste of time to rebuild everthing when a dependency changes, and sometimes it is the right thing to do, portmanager assumes it is always the right thing to do. One way this has proved to be a benefit is I've never had to run the special scripts when gnome is updated because after running portmanager everything is already up to date. Interesting. While I certainly don't mind a tool doing what's right, this issue which also exists with NetBSD's pkg_chk is the primary reason why I'm almost about to give up on it; it's just feasable to perform full system upgrades properly. Having your primary workstation half unusable for three days while the whole universe is rebuilding is not very nice... One possible solution I have considered for pkg_chk that may also work for portmanager is to set up a build environment in a chroot where everything is properly upgraded. Either for building packages for all upgraded ports such that the ports installed on the real system can then be upgraded quickly using the packages; or alternatively by perhaps maintaining two separate target directories such that one is being used by normal applications while the other one is being built. One could then make the switch atomically by re-mounting /usr/local (or /usr/pkg in NetBSDs case). Is this even feasable? Is portmanager intended to fully replace portupgrade in the long run? If so I would, as a user, very much value being able to upgrade all ports without disabling the machine in question. As it stands now, I much prefer portupgrade to NetBSD's pkg_chk for exactly this reason, even if portupgrade requires manual tweaking sometimes. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: traffic volume monitoring - what program
I am looking at a new ISP that charges for a certain number gigabites of traffic. I have -no- idea what my traffic volume is. Can anyone recommend a good traffic volume checker in the ports? I recommend setting up snmpd + rrdtool (portinstall net/net-snmp net/rrdtool). It's much easier and cleaner to configure than mrtg (in my experience). If you're interested I have a few scripts that do what you want in a basic bandwidth graphing situation for a few hosts; generating relatively sane graphs. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: darcs and ghc building
Does anyone know what is wrong with the ghc port? When trying to build darcs, I get: === darcs-0.9.23 depends on executable: ghc - not found ===Verifying install for ghc in /usr/ports/lang/ghc === ghc-6.2.1 is marked as broken: Does not compile on FreeBSD = 5.x. Without doing any investigation, I'll say that i recently installed darcs and ghc on a 5.3 system without difficulties, so it seems weird that it would be disabled. However it *WAS* a problem trying to get it to install on 5.2 due to the ghc bootstrap binary depending on a version of a library that is not in 5.2. Either something has happened which means ghc is no longer working, or perhaps it was incorrectly marked as broken due to the issue on versions below 5.3? Though - I just noticed you are installing 6.2.1. 6.2.2 is the one I have installed. Have you tried updating your ports tree? Perhaps the issue, whatever it is, is fixed in 6.2.2. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: migrating from thunderbird to mutt?
How do I get my old email (from thunderbird, mbox?) into mutt? mutt supports mbox/maildir/etc, so if youre using that you should be able to just point mutt to your mail. But then I dunno - perhaps thunderbird uses its own format? I got sick and tired of this issue years ago and now I read all my mail via IMAP, so that I know I am never *ever* dependent on one specific client. Can mutt handle um like 5+ email address and have them all separated and be able to send from diffrent email accounts? I dont know if mutt in and of itself supports that; when I need multiple personalities i tend to use different startup scripts for mutt that change the settings in question. Though you can surely create keyboard macros that change identities by changing relevant parameters. If I remember right mutt is just a mail reader, so how do I get mail to and sent from mutt? Correct. If youre using IMAP, you're done. If you're using pop, I think I remember reading mutt had some sort of support for POP. But in general, 'fetchmail' is probably what you want for downloading mail from remote mailboxes and having it delivered locally using whatever mail delivery mechanism exists on your system. mutt sends mail using the standard sendmail method; so having any properly configured MUA on the client should do (Postfix, qmail, etc...). How does it handle hyperlinks, if I select something will it open up in firefox or whatever? That sounds more like a terminal issue. Message filtering, for example I have all the different freebsd mailing lists automatically put into different folders, and junk mail sorting? Use maildrop or procmail for that (procmail is probably more widely used because its the traditional tool; maildrop I prefer personally because of its nicer syntax). maildrop/procmail is plugged into the normal delivery mechanisms, independent of mutt. use mutt to read the already filtered email. Speell check? Dunno; but i doubt it if you are talking about on-the-fly spell checking as you write. How do I set the thing up, config file help? www.mutt.org, and/or read through the example muttrc that comes with the freebsd port. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Good newsreaders for FreeBSD?
I'm looking for a newsreader that has multi-server capabilities in that it can piece together articles using different newsgroup servers. Similar to NewsPro for windoze. Anyone had any luck with a good newreader port for FreeBSD? I'm running FreeBSD 4.10 if that makes any difference. It doesn't have that particular feature you mentioned, but the one news reader that is closest to being perfect for me is slrn - very nice. As someone mentioned; leafnode can be used as a suitable proxy NNTP server that can pull news from multiple servers. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: vinum
drive a device /dev/ad0 drive b device /dev/ad3 ... ** 1 Can't initialize drive a: Operation not supported by device ... illuminating. Is this a common problem? How do I fix it? I've made sure that the disks in question have been labeled using disklabel -e as vinum volumes. What else? I suspect the problem is that you are supposed to specify the vinum volume; not the drive. That is, a vinum drive is a vinum partition on a physical device, not the physical device itself. So try specifying /dev/ad0s1a or /dev/ad0a or whatever is the name of your vinum labled partition. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Root fs full - free space always below 0
Hello, so during a portupgrade on my laptop the root fs, with soft updates enabled, became full. So I removed a bunch of stuff to make a few gigs available. I checked and df reported more than a gig of free space - so I re-ran portupgrade. Then I noticed it was full again, with df showing a negative amount of free space. I removed even more stuff, and rebooted just incase there were more blocks to be freed. After the reboot df showed a negative amount of space again. So I removed even more data (rm -rf /usr/ports/distfiles) and now I had 115 meg free df claimed. I then re-ran df in quick succession a few times and watched diskspace rapidly decrease to a negative 600 meg or so (note: the decrease was perhaps 150 meg/second, so it cannot have been a process writing data to disk in the background). After a couple more reboots and a manual fsck in single user mode I still have the same problem (on both CURRENT and 5.2.1-RELEASE kernels). What to do? -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: USB Storage devices (Cigar/Pen) in FreeBSD?
Hello, I'd like to buy one of those USB storage devices that come in different aspects (cigar/pen) and sizes (64/128/256/512 MB). My main machine does not have USB 2.0 ports (but does have USB 1) and is running FreeBSD-4.9 Do these devices work with FreeBSD-4.9? What kernel options should I enable? FreeBSD does have support for USB mass storage, so assuming it's a standard USB mass storage device, it should work. The biggest hassle in my experience is choosing a USB2 controller that works, if you want USB 2.0. On my laptop I have an intel controller, 82801DB/DBM (ICH4/M) according to pciconf -v -l, and it does work with USB 2 devices on CURRENT. However on 5.2 and below the controller dies when I insert a USB 2 device; though USB 1 works fine. Before the controller started working on CURRENT I went ahead and bought some random USB 2 controller for a stationary machine, since I was unable to find information on what works and what doesn't, but it turned out to not work either (in that I get 'weird' behavior in general; some times devices aren't detected and sometimes I can even get panics). This is on 5.2 though, I haven't tried CURRENT on that machine. It's a Via VT6202 based controller. So in short I think your best bet is to try to make sure you have a controller (built-in on the motherboard or otherwise) that is supported. Unfortunately I can't tell you what works and what doesn't, since I was unable to find that out myself. if you want to boot from the device. Otherwise I've not found anything interesting. Perhaps this is just because they work 'out of the box' and there are no quirks or issues that merit attention. I don't know about booting off them though. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: startx kde
on bsd i get an xterm the have to type startkde after startx is there a way to do it in one command? Create (or modify) ~/.xinitrc When X starts, it will execute ~/.xinitrc if it exists. If it does, X will also terminate when that file terminates. So simply putting startkde in ~/.xinitrc should do the trick. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: The journalling file system saga
Hello, I had to build a storage system this week with a capacity of 1.6TB. Regrettfully I decided to use Linux with XFS as the thought of waiting for fsck to complete in the event of a problem makes me wince. I experimented with FreeBSD, using two 800GB partitions and things like that, but in the end it comes back to the fsck if for any reason the machine goes down uncleanly. I share your reaction to the thought of fsck-after-crash, though I have come to appreciate softupdates lately after an obscene amount of googling. IMO the primary advantage to soft updates compared to journaling is that it allows good performance without write caching, since write operations can be deferred. The good part about this is that one can achieve good performance with write caching disabled on the drive/RAID, while journaling will be either slower with write caching turned off, or unsafe with it turned on. The question is whether that applies to data aswell as meta-data. I have not yet found any information as to whether soft updates guarantees the order of non-meta data (or: Is it safe to run PostgreSQL with soft updates?). If anyone reading this has a clue, I'd love to hear it. Unfortunately there are problems with soft updates, for me as a user. One problem is degraded performance with bgfsck, that you have already mentioned. Another problem is that bgfsck seems to be unsupported on the root filesystem (something which I am trying to fix, but it's going slowly due to lack of knowledge of FreeBSD aswell as lack of time). Yet another problem is that an fsync() no longer guarantees that data is on disk, even with write caching disabled on the media. This doesn't break things like PostgreSQL provided that the order of writes is preserved, but it does break things like MTA:s that want to guarantee that critical data has been commited to persistent storage before signaling success to an external entity (SMTP client). A very big issue is that soft updates addresses multiple problems - but it's an all-or-nothing choice. I can get good performance running safely (in some circumstances) by using soft updates, but if I need safety for an MTA I need to turn it off. But turning soft updates off does not only have the effect of decreasing performance, it *ALSO* creates the need for a full fsck after an unclean shutdown. But what if I need safety *AND* do not wish to have a 30 minute boot-up time? (Or in your case with 1.6 TB, I would imagine that's a LOT more than just 30 minutes...) A good solution might be to support *both* some kind of journaling/logging and soft updates. But to me that is still just a work-around for a broken foundation. I believe the fundamental problem lies in the ambiguity of fsync(). The same syscall is used to achieve different effects. A database like PostgreSQL with write-ahead logging (WAL) is concerned with making sure certain data is written before additional modifications are made (though see below). So it uses fsynch() to make sure everything is written before proceeding - thus causing a degredation in performance. But then comes qmail which needs to guarantee the data in question is *on disk*, and also uses fsynch(). This time the intended effect is specifically the goal of synch(). In the former case the intended effect was an implicit side-effect. PostgreSQL can be honored in terms of avoiding corruption (but not in terms of guaranteeing a transaction is commited to persistent storage when it returns) by softupdates provided that both meta-data and all other data is guaranteed to be written in the correct order (though again I don't know if this is the case). But qmail is not served by this. A filesystem that fulfills the requirements of qmail would also fulfill the requirements of PostgreSQL - but it would also unnecessarily decrease performance. Is anyone remotely interested in this? Yes, for the reasons mentioned below, and strictly for practical personal use because I'd love to be able to share data between FreeBSD and Linux ;) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: The journalling file system saga
[Problems with softupdates] Yet another problem is that an fsync() no longer guarantees that data is on disk, even with write caching disabled on the media. This doesn't break things like PostgreSQL provided that the order of writes is preserved, but it does break things like MTA:s that want to guarantee that critical data has been commited to persistent storage before signaling success to an external entity (SMTP client). fsync(2) works as advertised with softupdates enabled so that shouldn't cause any problems for applications that are written properly. Ah thanks! I was under the distinct impression that was not the case. This is indeed good news. I can now rest easy running PostgreSQL without disabling soft updates. Perhaps a document would be in order describing/detailing all the details such as this about soft updates, all in one place. What I know I have mostly gathered by reading papers and random mailinglist postings about certain specifics. Is there such a thing in existence? If not I may try to put one together, for other people like me. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: The journalling file system saga
Yes, for the reasons mentioned below, and strictly for practical personal use because I'd love to be able to share data between FreeBSD and Linux ;) Right now, FBSD offers the option to mount ext2 if you've compiled that into the kernel - I'd be happy to see a reiserfs option as well. If nothing else it would very useful to have the ability to mount reiserfs partitions under FreeBSD so that I could read the data I have stored there. Yes, I am aware of the ext2fs support. However I have avoided ext2fs like the plague for quite some time, and probably would not use it even to enable linux/freebsd file sharing... The amount of file system korruption experienced in the event of an unclean shutdown is just too great, even on a personal non-server system. So yeah, ReiserFS would be great. But I suspect it would also be a lot of work. I have googled on this in the past and it seems Namesys/Hans Reiser would be perfectly willing to implement such a thing if somebody were willing to pay for it. I personally think ReiserFS would be a very good choice, if additional filesystems were to be supported. I say this mostly in light of all the 'next generation' features of ReiserFS 4, rather than because of any actual or fictional advantages in performance of ReiserFS 3.x as compared to the alternatives. The idea of ACID transactions at the filesystem level is just too attractive to resist. :) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Is there any hardware RAID (SCSI) that is fully supported?
Hello, is there any hardware SCSI RAID controller that is fully supported in FreeBSD? By fully supported I mean being able to monitor and talk to the controller on a live system in order to initiate a rebuild on a replace drive and such. Mylex/Adaptec seems to be a dead end, LSI's MegaRAIDs I'm not sure about. Anything else out there? -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Mylex RAID management on FreeBSD?
Hello, I haven't been able to find much up-to-date info with Google on this. Is there currently a way to manage a Mylex AcceleRAID in FreeBSD? We are currently looking at either using an Adaptec RAID card or a Mylex AcceleRAID. Previous experience with Mylex has been good, but we have not used one on FreeBSD - only Linux. Are Mylex cards stable under FreeBSD? Any opinions would be welcome. Thanks! -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Vinum: How to safely remove a volume/drive from the configuration
Hello, so I had set up a rudimentary vinum configuration. I eventually rebooted the machine to make sure everything came up again correctly. It did, but it also picked up a couple of old drives and volumes that I had previously used for testing. vinum l now yields: 3 drives: D alpha State: up /dev/da0a A: 17244/488 MB (3532%) D backup_a_1State: up /dev/ad1s1b A: 31/19531 MB (0%) D backup_a_2State: up /dev/ad0s1a A: 39/19539 MB (0%) D beta State: referenced unknown A: 0/0 MB 3 volumes: V test State: up Plexes: 2 Size: 10 MB V test3 State: up Plexes: 1 Size: 20 MB V backup_a State: up Plexes: 2 Size: 19 GB 5 plexes: P test.p0 C State: up Subdisks: 1 Size: 10 MB P test.p1 C State: faulty Subdisks: 1 Size: 10 MB P test3.p0C State: up Subdisks: 1 Size: 20 MB P backup_a.p0 C State: up Subdisks: 1 Size: 19 GB P backup_a.p1 C State: up Subdisks: 1 Size: 19 GB 5 subdisks: S test2.p0.s0 State: up D: alphaSize: 10 MB S test2.p1.s0 State: crashed D: beta Size: 10 MB S test3.p0.s0 State: up D: alphaSize: 20 MB S backup_a.p0.s0State: up D: backup_a_1 Size: 19 GB S backup_a.p1.s0State: up D: backup_a_2 Size: 19 GB The drives 'alpha' and 'beta', along with volumes 'test' and 'test3' are old remnants. The question now is - how do I correctly and safely purge them from my vinum configuration? I tried: scode-whitestar# vinum stop test scode-whitestar# vinum stop test3 scode-whitestar# vinum rm -r test Can't remove test: Device busy (16) As I interpret the manpage, the use of '-f' should not be required. And indeed, I assume that '-f' should never be needed unless in extreme cases? In any case. At the moment my vinum configuration *is* expendable since I have not stored any data on any of the volumes yet, however I want to know how to do this correctly so that I can do it on a live system in the future. Btw, after 'vinum stop', the volumes state show as 'down'. But after an attempted rm -r, they go back to state 'up' (both test and test3, even though I only try to remove test). -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Nvidia Drivers
I have Nvidia Ge Force 4 MX 440. How should I install drivers for Nvidia Ge Force on FreeBSD 5.1 because FreeBSD won't work untill I install them? Well. I haven't used the nvidia driver on FreeBSD yet, but there is a port, so: cd /usr/ports/x11/nvidia-driver make install should do the trick; and follow the instructions in /usr/ports/x11/nvidia-driver/pkg-message -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Vinum: How to safely remove a volume/drive from the configuration
As I interpret the manpage, the use of '-f' should not be required. And indeed, I assume that '-f' should never be needed unless in extreme cases? In this case, you need the -f flag. The documentation is a bit fuzzy. Thanks! What is it about the situation that means -f is needed? In any case, it seems -f does not help. I did a stop on the test/test3 volumes followed by an 'rm -r -f test'. I received no error message, but 'vinum l' still lists the test volume - except now it is still in state 'down', rather than having reverted to 'up'. I also tried explicitly stopping the plexes/subdisks, but got 'device busy' (on the plexes/sds). I have included the full output of 'vinum l' below as it shows up now, in case I have missed something. Btw, I should mention that for reasons having to do with my previous testing/playing with vinum, the plex/subdisk relationsships are not what one might think by inspecting the output of 'vinium l'. Here's the output of 'vinum printconfig' (plus manual line wrapping to make it more readable): drive alpha device /dev/da0a drive backup_a_1 device /dev/ad1s1b drive backup_a_2 device /dev/ad0s1a drive beta device unknown volume test volume test3 volume backup_a plex name test.p0 org concat vol test plex name test.p1 org concat vol test plex name test3.p0 org concat vol test3 plex name backup_a.p0 org concat vol backup_a plex name backup_a.p1 org concat vol backup_a sd name test2.p0.s0 drive alpha plex test.p0 len 20480s driveoffset 819465s plexoffset 0s sd name test2.p1.s0 drive beta plex test.p1 len 20480s driveoffset 819465s plexoffset 0s sd name test3.p0.s0 drive alpha plex test3.p0 len 40960s driveoffset 839945s plexoffset 0s sd name backup_a.p0.s0 drive backup_a_1 plex backup_a.p0 len 39936000s driveoffset 265s plexoffset 0s sd name backup_a.p1.s0 drive backup_a_2 plex backup_a.p1 len 39936000s driveoffset 265s plexoffset 0s And 'vinum l': 3 drives: D alpha State: up /dev/da0a A: 17244/488 MB (3532%) D backup_a_1State: up /dev/ad1s1b A: 31/19531 MB (0%) D backup_a_2State: up /dev/ad0s1a A: 39/19539 MB (0%) D beta State: referenced unknown A: 0/0 MB 3 volumes: V test State: down Plexes: 2 Size: 10 MB V test3 State: down Plexes: 1 Size: 20 MB V backup_a State: up Plexes: 2 Size: 19 GB 5 plexes: P test.p0 C State: up Subdisks: 1 Size: 10 MB P test.p1 C State: faulty Subdisks: 1 Size: 10 MB P test3.p0C State: up Subdisks: 1 Size: 20 MB P backup_a.p0 C State: up Subdisks: 1 Size: 19 GB P backup_a.p1 C State: up Subdisks: 1 Size: 19 GB 5 subdisks: S test2.p0.s0 State: up D: alphaSize: 10 MB S test2.p1.s0 State: crashed D: beta Size: 10 MB S test3.p0.s0 State: up D: alphaSize: 20 MB S backup_a.p0.s0State: up D: backup_a_1 Size: 19 GB S backup_a.p1.s0State: up D: backup_a_2 Size: 19 GB -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Internet 2
Hello, DestGatewayFlags Refs Use Netif Razor 12.103.21.1 UGSc 2 105rl0 12.103.21/24 link#1 UC 1 0rl0 12.103.21.1 link#1 UHLW 3 2rl0 localhost localhost UH 0 0lo0 192.168.1link#2 UC 1 0dc0 kitty.my.domain00:06:5b:b4:41:1c UHLW 0 0 dc0 I don't see a default route anywhere in there. Add one with 'route add default IP_OF_THE_GATEWAY. Without a default gateway you won't be able to reach anything for which there is not a specific entry in the routing table. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: spam in an inbox.
How can I do that? Of course, I can make a program that decides where each message starts and where it ends, save it in a file and then filter it with spamassassin and with the filtered file use grep to find X-Spam-flag: YES to discard this message. I think it is too complicated... Is there an easier solution? If you can find a mailbox format converter (I'm sure there are a few but I haven't checked), you could convert it to Maildir and easily script the operation as with a Maildir each message will be a separate file. That way you don't have to do the parsing. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommended USB 2.0 controller fr. 5.2+
To be more specific I found a controllre by Q-Tec (425U) wtih a Via VT6202 chipset. Anyone know if this will work? As a follow-up: That Q-Tec controller (that I ended up buying) didn't work... USB1 works as usual, but no luck with ehci. When plugging in a usb2 drive either nothing happens, or I get an error followed by a kernel panic when unplugging it. (I posted already about this, offering help, to the usb-dev list. If someone here wants details I can provide them, but otherwise the above is mostly meant as a heads-up for future USB controller buyers reading the archives.) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Java runtime?
Coming from the windows world I'm used to having the Sun java run-time. I went to install it from ports, naturally...and can't figure out what I should be installing! Can someone give me a clue? Your best bet is probably java/jdk14 which is the native version of JDK 1.4.2. It will need to first install the linux-JDK though for bootstrapping purposes. Choose the JDK you would like to install and do 'make install'. It's going to ask you to download files from Sun manually due to licensing restrictions. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Java runtime?
Alright. I tried, and after hours and hours of compiling...this: Java HotSpot(TM) Client VM warning: Can't detect initial thread stack location /usr/ports/java/jdk14/work/control/build/bsd-i586/gensrc/java/util/CurrencyData.java:1: 'class' or 'interface' expected I haven't encoutered this particular problem myself, though I remember reading a post on this or the -java mailinglist not long ago where someone had the same problem. I did some google and didn't find anything conclusive, though I did find someone having the same problem on Linux when running in a chrooted environment. A suggestion which may or may not be correct is that it may be caused by the linprocfs filesystem not being mounted (because it broke in chroot). Try doing, as root: kldload linprocfs mount linprocfs /compat/linux/proc But again I have personally not seen this problem. When I failed to have a proper /proc filesystem for the Linux compatibility the result was the JVM busy-waiting indefinitely rather than getting instability/errors. But try that and see. If it doesn't fix it, write back. (I am assuming it is the Linux JVM, running under emulation, which is failing, given that it is compiling the standard library for the native JVM.) -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: disconnecting keyboard: big trouble !?!
Hello, sorry for butting in late here, but this keyboard plugging issue has been a pet peeve of mine for quite a while. But of course, the best solution to this whole hot-plugging issue is this: BUY ANOTHER KEYBOARD OR MOUSE. What you would rather do? Buy a $20 keyboard/mouse or a $150+ motherboard? Heck, you can buy cheap 4-port KVMs for under $200 these days too. That argument just doesn't hold in home environments or low-budget colo environments. Myself and pretty much everyone I know have a habit of unplugging and plugging keyboards freely and have done so for years and a multitude of machines, and so far I have never ever heard of it causing hardware problems. The reality is that it is *extremely* annoying when you need physical console access and find yourself dead in the water because apparently the keyboard was disconnected during boot. Whether that is because of some cabling issue, a lack of keyboards, or something else isn't important - it can easily happen at home. And it may very well lead to some database corruption (or what-have-you) if you are forced to do an unclean shutdown. In a colo situation it can be much worse. In order to remedy the situation you need a digital KVM switch since an analog switch is essentially emulating plugging/unplugging quickly rather than acting as an active keyboard for each machine it is connected to. Not only can such a switch cost more than $200 (at least the ones I've seen) - they can also break (which happened to the one and only digital KVM switch I have ever used), and they waste precious rack space. And even with a KVM switch the flexibility of being able to re-arrange cabling is quite useful. Serial redirection is not always an option in a low-budget environment such as personal colo boxes where you try to really press the costs. How many low-budget motherboards have you seen that support console redirection? Not many I would guess. In short - it may seem like a trival problem but it's helluva annoying and can easily lead to serious problems both at home and in colocation environments. I'm very surprised FreeBSD doesn't handle this by default out of the box. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: disconnecting keyboard: big trouble !?!
On Tuesday 23 March 2004 19.44, Matthew Emmerton wrote: That argument just doesn't hold in home environments or low-budget colo environments. If you're low-budget, buy another keyboard! And have ten+ keyboards in the rack? At home? It's a pain. And even with ten keyboards it's a pain having to reboot a machine just because the keyboard died or because the cabling needs to be re-arranged. FreeBSD should not be working around user's bad practices. It has *always* been a bad idea to hot-plug PS/2 peripherals, and until USB is the norm, will continue to be the case. Again I've never heard of anyone running into any kind of problems due to this. And someone else recently posted about not being able to find any conclusive information on this info at all while googling. So really I don't consider it bad practice at all, since I have never heard of a failure, nor seen any concrete evidence why it would actually be bad. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Recommended USB 2.0 controller fr. 5.2+
Hello, I am planning on buying a PCI USB 2.0 controller for a machine running FreeBSD. The only experience I have so far is with my laptop which has some Intel controller that there are problems with, leading to controller shutdown when plugging in a USB 2.0 mass storage device. I read somewhere that there was a problem with that particular controller but I don't remember why. So - if I want a USB 2 controller that works fine with FreeBSD, which one shojuld I get / which chipset should it be using? Thanks! -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommended USB 2.0 controller fr. 5.2+
So - if I want a USB 2 controller that works fine with FreeBSD, which one shojuld I get / which chipset should it be using? To be more specific I found a controllre by Q-Tec (425U) wtih a Via VT6202 chipset. Anyone know if this will work? -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Rxvt replacement?
Is there any other xterm replacement which is small like rxvt? Or at least smaller than xterm. Grepping around the ports tree i found 'wterm' which is supposedly a fork/branch of rxvt (haven't tried it yet). There's also aterm (also forked off rxvt apparantly). -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD 5.2.1 and Bluetooth
We will shortly be starting some development with bluetooth and I'm looking to setting up a workstation dedicated for that. This workstation will also be used for palm os development since that is part of the same project so the workstation needs to be somewhat stable for that. Is the bluetooth support in FreeBSD 5.x well developed? The main protocols we need are RFCOMM and HID support. I've heard some problems with FreeBSD 5.2 as far as stability, but are they mostly fixed in FreeBSD 5.2.1 or should I just go with Debian Linux for all of this. FWIW I have successfully connected to the internet via GPRS over bluetooth with a Nokia 7650 and the built-in USB adapter in the IBM T40p. It seems to run stable (well except for the buggy-in-general phone). Dunno about HID though. -- / Peter Schuller, InfiDyne Technologies HB PGP userID: 0xE9758B7D or 'Peter Schuller [EMAIL PROTECTED]' Key retrieval: Send an E-Mail to [EMAIL PROTECTED] E-Mail: [EMAIL PROTECTED] Web: http://www.scode.org ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]