On Thu, 3 Apr 2003, Stephen J. Turnbull wrote:
> Er, now that rings a bell. Is everybody aware that much djb code
> suffers from legal problems that potentially make it unmaintainable by
> Which of the several djb (non-)licenses apply here?
None.
rwcdb is written by Jan from scratch.
Cheers,
Hello Greg,
nice to see a detailed analysis, even if I argue against some points.
> When one modifies the kernel, and changes a structure definition that
> affects an interface, installing the kernel needs to install the new
> .h so that programs can be rebuilt for the new interface.
This interf
Hello Greg,
On 2 Apr 2003, Greg Troxel wrote:
> NetBSD and FreeBSD have db 1.85 as part of the base system.
> ... gdbm ... required by ...
> ... sleepycat db3 installed, required by ...
> So while a rwcdb implementation is nice, using native db, gdbm, or db3
hmm, may be not all systems are as
Hello Stephen,
> Er ... has libdb disappeared from the most recent Coda CVS? I can't
it has been replaced by rwcdb (Jan's from-scratch implementation of
read-write interface to cdb). It feels so much better since then.
> log in and ldd (this is a Debian Linux system) says libdb is not linked.
On Mon, 31 Mar 2003, Simone Sestini wrote:
> I'm installing a CODA farm for a mail project..
> Today i'm configuring the SCM Server and i did that kind of partition.
>
> Device BootStart EndBlocks Id System
> /dev/sda1 164514048+ 82 Linux swap
> /dev
On Fri, 28 Mar 2003, Greg Troxel wrote:
> Coda _has_ worked fine on BSD for a long time. There was a change in
Sure.
I was thinking about the realms-aware code, as I have had problems on
FreeBSD.
> NetBSD vfs locking rules for lookup recently (in 1.6 but not 1.5), and
> that introduced a probl
Hello Greg,
you did a good work!
> This is hard to follow if you don't grok vfs locking rules, but FYI
> for those that do. (I'm only newly almost in that camp...)
I do not really belong there, but I am glad there is progress on *BSD,
partly as we have FreeBSD machines that might become casual
Hello Steve,
pleasant to hear that you are taking the Coda evaluation orderly.
> the problem i'm having is that write operations are painfully slow.
As Jan has described, it is the synchronous nature of the operations on
metadata that makes updates inevitably slow, in strongly connected mode.
Da
Hello Sandeep,
> Upon network partition healing, do replicated volumes automatically
> synchronize all their objects, or are objects synchronized upon next
> reference? So if f is replicated and I modify f in one volume because
the replication is "lazy", so that you have to run at least a stat()
Hello Simone,
> Today I'm planning to start the new coda server for a mail project.
>
> I have here 3 machine, with a 73gb u160 disk.
your setup is going to be the largest known (at least to me :)
> How is the best way to do partition and all ?
On each host:
Create a file data partition for ea
On Sun, 16 Mar 2003, Liu Kang wrote:
> Maybe I misunderstood you, do you mean put a PROM in coda-server?
No. I was talking about clients.
Regards,
--
Ivan
Hello Liu Kang,
you write:
> but have you considered about the cost? The method you provided may be
> used in commercial distribute storage system, coda is an open source
Unfortunately I do not understand what you mean.
The idea was not really about distributed storage, wasn't it?
> software. I
Hello,
I have figured out a couple of places where Coda can make life so much
easier.
One "hot area" - thin clients.
--
I have looked at open source solutions.
Still the most popular ones are X-terminals on Linux/*BSD kernel on
NFS-root or tftp-able RAM disk.
(very ins
Hello Jan,
> On Sat, Mar 15, 2003 at 05:30:56PM +0100, Ivan Popov wrote:
> > and that we still lack multilevel backup volumes (despite that they would
> > fit very naturally, but also afs and dfs people missed it).
>
> Ehh, we have had multi-level backups working reliably for
Hello,
(the following is not in any way promotion for any commercial product,
in case you might have a doubt)
yesterday I have been at an informational meeting SUN Microsystems
arranged here. They are very proud of their backup management product,
SAM-FS:
"... a scalable, flexible storage with su
On Wed, 12 Mar 2003 [EMAIL PROTECTED] wrote:
> Does coda run on 64 bit solaris ?
Hello Jane,
unfortunately the answer for the moment is "no".
It works with 32 bit kernels only. There was (an attempt of?) a
version of the kernel module for 64-bit Solaris, but it was for a long
time ago and nobod
On Sun, 9 Mar 2003, Jan Harkes wrote:
> > So my question is - how much would have to be rewritten to be able to use
> > some other database, allowing the same persistency and resilience,
> Not sure how much would have to be rewritten, but a lot (most) of the
> server code isn't really using RVM a
Hello,
I am doing "brain surgery", as Jan calls it, on my Coda
servers - moving and resizing the RVM storage.
The process takes some time and gives an occation for thinking.
RVM is probably very efficient at the access, as all data resides in
(virtual) memory when you need it.
On the other side
On Sun, 9 Mar 2003, Liu Kang wrote:
> Why not set up a cvsweb interface like cvsweb.freebsd.org and many other
> program?
You may have missed
http://www.coda.cs.cmu.edu/cgi-bin/viewcvs.cgi
Cheers,
--
Ivan
or immediate migration to Coda.
As I do not know your usage pattern and the goals of possible migration, I
cannot guess more.
> - Original Message -
> From: "Ivan Popov" <[EMAIL PROTECTED]>
> To: "David Shirley" <[EMAIL PROTECTED]>
Hmm, do you rea
Hello David,
On Mon, 3 Mar 2003, David Shirley wrote:
> We are considering switching from NFS to Coda, in our
> current setup we have a 270 Gig RAID 5 paritition.
>
> My quesiton is do we really need to have 4% RVM partition?
> Also must this be on a seperate disk or can it be taken out
> of the
Hello,
I want to share a success story, and document the steps I did
(in essence, followed Jan's recommendations and read some man pages)
Running a two-server setup with all volumes replicated.
One of the servers got a corrupted volume and refused to start,
complaining about an assertion error w
On Fri, 28 Feb 2003, Brian White wrote:
> > If the 'observed' bandwidth of the link to the servers A,B, and C is
> > above the weak-connected threshold (about 64kB/s) the Coda client pushes
> > updates to all servers.
>
> If there are many sites, does only one have to be belowe the threshold
> bef
On 27 Feb 2003, Steffen Neumann wrote:
> Your sysadmin will have different things in mind
> if his SCM (very often a central machine) is down,
> and delay other things anyway ;-)
:-)
I thought about something like planned maintainance,
otherwise of course, SCM usually hosts some volumes too
so y
Hello Eser,
On Thu, 27 Feb 2003, Eser Utku wrote:
> I was wondering whether there is any possibility of having full redundancy
> with Coda since, from what I have gathered, when the SCM server is
> disconnected or crashes the clients can`t reach shared files that are
> created after the SCM is ou
> > What does "ls -l .openoffice/1.0.1/user/temp/soffice.tmp/sv1o9.tmp"
> > (one level up) show?
> since sv1o9.tmp is a directory, I get the same error as above, with
> option -a it correctly lists the . and .. entries but still complains
> about sv1oc.tmp. ls -l soffice.tmp correctly lists sv1o9.t
Hello Jochen,
> $ find /coda/usr/coffee -type l -exec ls -l {} ;
>
> doesn't show any objects in conflict (strange enough, find /coda/usr
> doesn't recurse into coffee/...)
at least gnu find has some "optimization" that is based on assumptions
wrong for Coda (see find's docs if you are curious).
On Mon, 17 Feb 2003, myciel wrote:
> I was considering migration of my mail serwers from nfs to coda, but from
> last line of coda 5.3.20 Changelog
> I learned that there is file number limit ~24. Is overcoming of the
> limit really impossible?
It is the max number of files per volume, seem
Hello,
as I couldn't easily find the relevant reference, I post it here,
to be sure it is archived and available for somebody else:
Jan once described the steps necessary to recover rvm from "date had been
far in the future" situation.
I had to use one extra command, compared to the decription, I
> Coda clients are "server" that menage Imap and pop3 connection and write
> email over the coda server :)
Ok, one Coda client, multiple servers. I assume you choose Coda for the
storage failure redundancy and will replicate all data across Coda
servers.
You should build a test system and test it
On Tue, 4 Feb 2003, Simone Sestini wrote:
> Hi folks..
>
> I'm configuring a professional mail server using qmail and coda..
>
> Now i'm planning to use 3 coda server on my storage-farm, each one is an
> AMD XP 1800 with a GigaEth , 256 Mb of ddr and 73 Gb of U160 SCSI HD.
>
> Now i would to know
Hello!
Some of you can recall a posting called "design: go beyond AFS?"...
I spent several years using and adminstering "global file systems" like
DFS and Coda, have used AFS, too.
All of them share the concept of volumes/filesets and of "mountpoints" as
special file system objects.
A lot of ti
On 27 Jan 2003, Steffen Neumann wrote:
> > > or do you 2) *expect* them to be inconsistent simply because one
> > > was down for some time ?
> I think you don't trust coda enough that it could handle this
> situation itself, but I am pretty sure it does.
Sure it does. I have (inadvertently) been
On 23 Jan 2003, Vojtech Moravek wrote:
> When I have two servers is cell and non-SCM is down. Data on SCM server
> will be different after non-SCM start. My question is. How I can
> synchronize different data on volume between SCM and non SCM??? and is
> the same situation when SCM will be down??
On Mon, 13 Jan 2003, Jan Harkes wrote:
> Kerberos is something that I don't use myself, so it is mostly dependent
> on patches that people submit. The problem is that sometimes I get
> carried away and clean up the code a bit too much. I believe that Ivan
> Popov confirmed tha
Hello Jochen,
On Sat, 11 Jan 2003, Jochen Eisinger wrote:
> is it possible to automate the login process to a coda-server (using
> clog) a little? I'm thinking of a pam module or similar that obtains
> coda tokens when a user whos $HOME is codafs-mounted (assuming that both
> login and password a
On Mon, 30 Dec 2002, Jan Harkes wrote:
> > I am rebuilding lwp,rpc2,rvm libraries and I see that libtool is used and
> > that it tries hard to compile library paths into the libraries and
> > the binaries.
>
> Libtool only hardcodes paths when the destination path is not in the
> default run-time
Hello!
I am concerned about the default behaviour of one of the Coda building
tools, namely libtool.
I am rebuilding lwp,rpc2,rvm libraries and I see that libtool is used and
that it tries hard to compile library paths into the libraries and
the binaries.
While it *is* possible to live with that
Hello Greg,
On 19 Dec 2002, Greg Troxel wrote:
> Right now, you can run coda over IPsec to get some reasonable
> security, or over NAT. But you can't do both.
I would guess that making Coda more secure by implementing strong
encryption is as hard/easy as making it work over ipv6. Both approache
On Sun, 15 Dec 2002, Mark Dickey wrote:
> I don't want to be a pain, but if anyone even has a Yes/No answer as to
> whether or not I should investigate and test Coda further, I'd appreciate
> it.
Sorry for the silence Mark,
probably no one answered just because this topic has been discussed very
Hello Jan!
> Yes, but people who still think 'NFS' will be even more tempted to
> ignore that requirement, and probably either not read, or get confused
> by instructions that say 'do not modify any files in the exported fs'.
Alas, you are right.
> Needed, tar<->voldump conversion tool, and a vo
On Mon, 9 Dec 2002, Jan Harkes wrote:
> > It *could* be possible to rewrite the server (and probably have to
> > implement a local file system suitable for putting the files on it), so
> Not really possible. The current semantics is that a modification on the
> server is an atomic change from the
On Mon, 9 Dec 2002, Jan Harkes wrote:
> > exported by Coda. No offence meant. Coda offers very special properties
> > that are hardly synchronizable with other "access paths", i.e if you'd
> > insist on accessing your files via the local filesystem.
>
> Not only that, but the server wouldn't neces
Helo Josie,
> eventually gave up. So I decided to try again yesterday and I finally got
> it working... at least so it seems.
glad you did it!
> What's bothering me is that the files exist in coda land. It's not like a
> way of accessing files from other peoples computers.
No, and you are right
On Fri, 6 Dec 2002, Fernando Schapachnik wrote:
> Hi Ivan,
>
> Thanks for you answers. Let me refrase my original question: are there case
> studies/success stories of large organizations using CODA? This is for the boss :)
I think the answer is no
(any knowledgeable person please correct m
Hello Fernando,
> I'm evaluating CODA for usage in a large network and two questions:
>
> a) Is it usable in a 3500 workstation network?
I do not think somebody had tested Coda for similar setups.
Coda creates very modest traffic as long as the files can be
cached (i.e. unless there are mas
On Wed, 27 Nov 2002 [EMAIL PROTECTED] wrote:
> I read that it was recommended not to make a server larger than 3.3GB. I
> want to create a server that is around 60-65GB worth of data. Has anyone
> made larger partitions? Any performance penalties? Any recommendations?
Hello Matt,
this issue has
On Wed, 27 Nov 2002, matthewh wrote:
> Hello all,
>
> I am new to Networked file systems and I read about coda and I wanted to try
> it. I am trying to install from source on FreeBSD 4.7 (I also tried on a
> FreeBSD 4.5 machine and I have the same problem).
> Configure completes but when I run mak
Hello Anthony,
> quick question:
>
> I have successfully set up the server, created a root volume named
> "testfs". My question is, where do i put the files on the server so that
> they will be available in /coda on the client?
You do not put files "on the server", you put files via the clients,
On Wed, 20 Nov 2002, [windows-1252] Toma Ianc wrote:
> how to create new share or just link directory on server with coda ?
Hello Toma[?] ,
sorry for not writing your name correctly but combination of my terminal
and mail client do not fully support windows-1252 encoding.
> or better where to
On Thu, 14 Nov 2002, Timothy Murphy wrote:
> I may say I have no intention of using 2GB files,
> and am not too concerned about the speed.
> But I don't want all my files to disappear!
Hello Timothy!
I have probably different mental filter settings :-)
but I cannot remember people saying they ha
Hello Simone,
> /dev/sda1 20Gb /
> /dev/sda3 2Gb /vice
> /dev/sda4 500M RVM LOG Partition
> /dev/sda5 60Gb RVM Metadata
> /dev/sda6 1,6T /vicepa
> 2) The RVM Metadata partition should be 4% of the Storage Area so about 60G
> but howto say don't use more then 315M
you are limited by 1G (tested) o
On Mon, 11 Nov 2002, Jan Harkes wrote:
> On Mon, Nov 11, 2002 at 10:19:31AM +0100, Fabrizio Morbini wrote:
> > Hi, Can Coda handle file with size bigger than 2 Giga?
>
> I'm pretty sure it doesn't. And would even go as far as to question
> whether it is all that useful.
Hello Jan,
I think that i
Hello!
> The frontend servers will not have the same content. So even if
> user upload files through then, it will be no problem. The big
> opperation here will be the reading.
Sounds good. The need of personality switching still applies. You want to
let the web server process acquire Coda
Hello,
> I'm setting up a webserver farm with 3 machines. 2 will be frontends
> with apache and one will be the storage server.
>
> I was thinking about using NFS to do that, but maybe CODA will be a
> better solution?
it depends on your needs. NFS is surely easier to set up,
Coda has som
On Wed, 30 Oct 2002, Marcel Pol wrote:
> Hello,
> I tried to get coda from today's cvs built, but it wouldn't build a functional
> configure file.
> I'm not sure what happened, but part of the configure file is missing.
> I don't have the expertise to fix it myself, so I hope I can give enough
> i
On Wed, 30 Oct 2002, Marcel Pol wrote:
> Ok thanks for the information.
> I'll be trying out a cvs version then, and see how it goes.
> I assume I only need to download coda from cvs, and not lwp, rpc2 and rvm,
> right?
I would recommend cvs rvm, too, as it fixes some potential errors.
Looks like
Hello Marcel,
On Wed, 30 Oct 2002, Marcel Pol wrote:
> Hello,
> I started using codafs very recently, and I noticed it only builds with
> gcc-2.9x.
I cannot give you any guarantee of course,
but I am running recent Coda (cvs, later than 5.3.19) and using gcc 3.2.
So far I have no good reason to
On Mon, 28 Oct 2002, Jan Harkes wrote:
> On Fri, Oct 25, 2002 at 05:12:46PM +1000, Timothy Kersten wrote:
> > open_log failed.
> > do_rvm_options failed
> > 17:09:14 rvm_init failed RVM_EIO
>
> Actually it is quite unusual that open_log fails. Are you sure rvm is
> initialized correctly?
We are
Hello Mark,
> I am trying to add public/private key authentication to Coda (Grid Security
> Infrastructure - www.globus.org/security), much the same way that kerberos
> support was added. I am quite unfamiliar with the source code and thought
a nice idea!
> it might be a good idea to ask where w
On Fri, 25 Oct 2002, Timothy Kersten wrote:
> This is what i get from "/vice/srv/SrvLog". Have tried re-instaling and
> followed all instructions. It's driving me nuts!
>
> 17:09:14 Server etext 0x80bfdba, edata 0x80ec210
> 17:09:14 RvmType is Rvm
> 17:09:14 Main process doing a LWP_Init()
> 17:09
On Thu, 24 Oct 2002, Nathan Ward wrote:
> damn, theres no way to remove this cache?
> I need a filesystem with the security/acls like coda, but with no limit
> on file sizes...
If you do not need resilience to server/network failures, you do not need
Coda.
OpenAFS is caching, but not the whole
Hello,
thinking of smooth ways to let a group of people create volumes
without distributing a "Coda super user" password.
One way is of course login-authorization on scm, letting people run
scripts as super-user (e.g. via sudo) and thoroughly checking their input
and arguments, with a homegrown "
On Mon, 21 Oct 2002, Jan Harkes wrote:
> primaryuser= option in venus.conf and if that doesn't exist whoever
> is logged in /dev/console or /dev/tty0 (or was that tty1?)
Ok.
> But a whole lot are probably not adequately secured. Commands like cfs
> disconnect and cfs reconnect really work at the
On Wed, 16 Oct 2002, Kevin Atkinson wrote:
> One Question though. Is the current implementation of AFS the same way?
> That is does AFS also insist on reading the entire file before any parts
> of it can be access? If so maybe I will look at AFS. However, AFS from
> my experience AFS is a pain
Hello Kevin,
> > But a close() after writing and open() before reading will be very
> > time-consuming. I haven't tried Coda on 100Mbit network but I can't
> > expect over 7Mbyte/sec anyway, so when you start a playback of a 2000M
> > file from Coda, you will have to wait 5 minutes at least (the
Hello Kevin,
to avoid going into all details I am trying to find the critical one(s):
I am a bit afraid you expect NFS- or SMBFS-like behaviour from Coda
concerning latency.
On Tue, 15 Oct 2002, Kevin Atkinson wrote:
> machine. Then copy the files which vary anywhere from (100 - 2000 MB)
> Co
Hello Gary,
> The one thing that I'm trying to do (which may exacerbate the problem
> is):
>
> * 'clog' as user X
> * perform a number of operations on files/directories, owned by X
> * 'cunlog'
as Greg pointed out it is going to create problems if you cannot guarantee
strong connecti
On Sun, 6 Oct 2002, Jan Harkes wrote:
> createvol_rep testvol verdi/vicepa mozart:/foo marais
>
> which would create replicas on the codasrv on verdi in partition
> /vicepa, on a server running on mozart (at port ) in partition foo,
> and on marais on the default partition.
Yeah!
> But
Hello!
Can somebody tell about his/her own experience or make a suggestion based
on the code knowledge -
how well do Coda groups (and e.g. the pdbtool utility) scale?
Is it ok to have thousands of users?
[assume yes except that the "list" command becomes impractical]
Is it ok to create a group
Hello,
I am a bit confused about the management of volume replicas.
Sorry to admit, I have not read the source to get the definite answer, but
it would be a rather time-consuming exercise. Hope somebody has the
answer at hand :)
Right now it is not supported (?) to modify a volume storage group
Hello Adrian,
> I'm going to be building a new, dedicated fileserver soon, which will
> feature a more formal arrangement shared across more servers and
> clients.
I would suggest you consider a volume naming model where a volume name is
just its mountpoint path. Volume names may include '/' so
Hello Adrian,
> Thanks for your reply; glad to have it all working again. Have upgraded
> both the server and all clients to 5.3.19.
nice to hear! I would suggest upgrading the kernel to 2.4.19 too,
I think kernels <2.4.19 include older Coda module code, that is fixed in
2.4.19.
Regards,
--
Iv
Hello Adrian,
> * I have a token;
> * the client can talk to the server (cmon, cfs cs);
> * there are no conflict-signifying dangling symlinks (find /coda
> -type l).
> The logs mention plenty of entries regarding how the volume has
> 'unrepaired local subtree(s)' and h
Hello Brian!
On Thu, 3 Oct 2002, Brian Widdas wrote:
> I'm a bit confused - if you have strong connections, and you're going to
> be updating the repository from several computers, doesn't this seem like
> you're trying to reinvent the wheel?
Hmm. Reinvent the wheel? I am trying to use a sunlig
Hello,
I have tried to look in the archives but that did not help much.
Does somebody have CVS repositories on Coda, and if so, how much disaster
cause possible file update conflicts? I mean accessing the repository via
the filesystem, not pserver or ssh. I mean also team-style updates, not by
o
Hello!
I have played with replication and failovers and it worked very pretty,
but I have noticed a strange thing that shouldn't have happened.
It was inconsistent user database on the slave server when I shut
down the scm and tested authentication.
Files *.db had rather old (and matching with t
On Thu, 26 Sep 2002, Jan Harkes wrote:
> > dsparm=1073741824
> > heapsize=0x3FF0
> > rvmstart=0x5000
> > staticsize=0x10
> > nlists=80
> > chunk=32
> But looking at your other numbers, heapsize = 0x3ff0 and static size is
> 0x10, which adds up to exactly dsparm (0x4000). But RVM need
Oops, I should have used search in the archives, I have evidently missed
that exactly this message has been discussed.
[then, I *know* how to spell "worry"... :]
On Thu, 26 Sep 2002, Ivan Popov wrote:
> release_segment unmap failed RVM_ENOT_MAPPED [this is what warries me]
It
Hello!
A host with 128M RAM, 1G partition as rvm data, 1.2G partition swap.
A brand new minimal Debian install, custom kernel 2.4.19.
At vice-setup time rdsinit tells:
(typing looking at another screen)
-
Rdsinit will initialize data and log.
T
On Thu, 26 Sep 2002, Timothy Kersten wrote:
> I still have no sollution. I have followed all the steps for
> installation but codasrv just vanishes? Thanks.
In a similar situation I have substituted codasrv start in the script by
strace . >/tmp/log 2>&1
and then looked at the last syscalls
On Tue, 24 Sep 2002, Ivan Popov wrote:
> I'm running typically with 90Mb cache.
Oops! No I am not. :-)
I mean 90 K == about 900 M of course.
Hope nobody has taken literally.
Yours,
--
Ivan
Hello Derek,
On Tue, 24 Sep 2002, Derek Simkowiak wrote:
> > Unfortunately, you can not count with having more than about 1G RVM per
> > one server process.
>
> First, can you substantiate your claim about the 1 Gig RVM limit?
> I just need something harder than "some guy on the mailing li
Hello Derek,
trying to answer some questions based on my own experience, your mileage
may vary.
On Tue, 24 Sep 2002, Derek Simkowiak wrote:
> load-balanced cluster. They will drag and drop files to this WebDAV
> share, which will result in Apache/mod_dav writing a file to the the CODA
> filesy
On Tue, 24 Sep 2002, Jan Harkes wrote:
> > On Tue, 24 Sep 2002, Ivan Popov wrote:
> > After consideration I want to withdraw this idea as it implies a need for
> > complementary unix-like uid-based access control.
> > My conclusion: PAGs are of no real use!
> opera
On Tue, 24 Sep 2002, Ivan Popov wrote:
> *** well, PAG would help to allow cron jobs to alter user files on Coda if
> the user explicitely grants host/ the right to do that...
> The same for mail delivery and other "problematic aspects" of networked
> filesystems.
After c
Hello,
a background: I have noticed somewhere in Coda code a special case for
uid 0's credentials (venus invalidating acls on cached objects, something
like that?).
I am a bit concerned about uid 0 being able to access /coda as an
authenticated user (thus protected from server spoofing), without
On Wed, 18 Sep 2002, Greg Troxel wrote:
> Yes, I have seen this sometimes. The repair just wedges or returns an
> error and after that the client seems to be a quivering wreck
> requiring an init to get it right.
>
> OK - so I am not the only one seeing this bug.
I suppose this is the pro
Hello Johan!
> that I'm currently workning on my M.Sc.-thesis at Chalmers University of
> Technology in Sweden and one of the issues that I am "investigating" is
> distributed file systems.
Nice to hear!
> Does anybody know of any (benchmark-)comparisons between NFS and Coda or
> AFS and Coda?
> On Thu, Aug 22, 2002 at 10:43:15AM +1000, Nick Andrew wrote:
> >
> > I guess it must be hard to do, if people always talk about it but
> > never do it. I imagine that the union filesystem code has to map the
In fact, it has been done many times, for Linux too. Then abandoned.
Now union mounts a
Hello Jan,
> > "special" paths at each mountpoint.
>
> I'm not sure I follow you here
Then forget it, see below.
> I don't want to make access to underlying replicas easier except during
> repair.
I see. Ok.
> In the end both the client and server still only think about volumes by
> their num
Hello Jan!
My vague idea about magical path beginnings to access underlying
volumes was not really though out, just seemed nice while I wrote.
As the decision about how to treat a certain mount point has to be
done for the mountpoint, not for the whole path (that may contain multiple
mountpoints
Hello Nick,
> I am in the process of setting up a _home_ fileserver with twin 80-gig
> disks (RAID-1 mirrored) and am looking for a distributed filesystem
It should be doable, depending on how big your files are - i.e. it is the
===> number of files <===
being the limitation, not the total siz
Hello Jan,
thanks for the analysis!
Glad it looked reasonable for you too.
On Tue, 20 Aug 2002, Jan Harkes wrote:
> I see a point for larger volume names, although that breaks the existing
> client-server rpc's. I think the '/' character can be used already.
Wonder if the constants are "hardc
Hello,
I have thought about experiences of AFS and DFS deployment at our
university since 1997, where I have been rather heavily involved,
and my own experiences with Coda.
I have not seen anything that would really need the separation between
mount point name space and volume name space.
Rathe
On Sat, 3 Aug 2002, Brett Lymn wrote:
> Why does Coda have to solve this at all? It sounds to me that what
> you are really wanting is an IPSec security association.
Coda is already solving a lot of security problems and has built in
means for doing that.
IPSec is a solution at a lower lever,
Hello!
I am thinking about server spoofing (an issue becoming more visible with
coming multi-realm architecture).
With authenticated access it is possible (unsure how much of it is
really done?) to check that the fileserver is who it pretends to be, as
authserver could prove that it knows the us
On 9 Jul 2002, Erik Enge wrote:
> Is there somebody out there using Coda in a production
> environment?
I think Carnegie Mellon is still the place where Coda is deployed mostly.
I am running 2 dedicated servers and 6 clients for 5 users running mostly
mail/irc/www with all of data occupying
Hello,
pity to hear that people are giving up with Coda.
The reason *might* be in trying to hide the complexity and "make it easy",
while creating excessive complexity for that purpose?
My own experience is that for such a complicated service as
coda server or client, it is very hard to make a
On Thu, 13 Jun 2002, Phil Nelson wrote:
> The current Solaris support was done under 2.6 and was 32 bit.
I can add that it worked (i.e. for me) in 32-bit mode on Solaris 7 too.
Regards,
--
Ivan
1 - 100 of 172 matches
Mail list logo
