On Mon, Feb 02, 2015 at 05:55:26PM +0100, Thomas HUMMEL wrote:
> Hello Timo,
Hello again. I'll try to answer my own question myself ;-)
I think my problem was that "pass=yes" just becomes _mandatory_ when using
_only one_ and _prefetched_ userdb because retrieving the
On Mon, Feb 02, 2015 at 05:55:26PM +0100, Thomas HUMMEL wrote:
> I just tried the master user feature with a very simple setup (Dovecot
> v2.2.15) :
Also, the documentation states that :
"pass=yes" doesn't work with "LDAP with auth_bind=yes, because both of them
t
supposed to be in the 'nonesxistent login username' case.
Can you help ?
thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
handle the
proxying of both IMAP and POP3 (especially for the same user who can use both)
?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
like this -
> - is to combine a ldap passdb with deny=yes.
I thought about that too, but that would mean setting up another LDAP
directory, which I find a little bit overkill.
Thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
, then access is granted ?
> b) or could I use only one ldap passdb by changing the pass_filter
>
> from
>
> pass_filter = (&(objectClass=posixAccount)(uid=%u))
>
> to something like
>
> pass_filter = (&(objectClass=posixAccount)(uid=%u)(!foo=yes))
>
me user in different
userdb's ?
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
nvolve this host and only this host...
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
atd and lockd restarted
and the filesystem un-and-re-mounted while the same setup (up to date via
FreeBSD update, i.e. same base, same nfs client) would not make the server log
the message.
Anyway, the problem seems harmless. But is it legit that dovecot try to unlock
non (or no more) locked files as it
not, noticing there isn't anything to
unlock.
Can you help me explaining such a behavior ? Are those "unlock a file with no
prior lock" made on purpose or is it a bug ? Would it be an application or a
RPC bug ?
Can you think of another reason ?
Thanks
--
Thomas Hummel | I
ct. But
as you said dsync knows how to handle this and would assign new uids to both
and the client, as you said, could get confused about what he thought was 123.
But even in that case, wouldn't he see the message (as a new one with it's new
uid) ?
I mean, nothing would be "lost&
obaz comes in. Oh, I get it, you mean since
uids gets only incremented, this new could not get uid 100 and then confuse the
client index ?
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
r deletes a mail foobar in the new server because he
reconnects
after the kick. I guess dsync would merge the change and would not sync the
foobar message from the old server in the final step. But what if another ,
new, mail foobaz is
delivered : would'it get the nextuid which was the uid of
est than rsync (maybe because using
dovecot-uidlist or similar) ?
Besides, how about client side indexing while in this incoherent, not yet fully
sync'ed state ? Wouldn't there be corruption risk ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
On Thu, Apr 11, 2013 at 01:09:18PM +0300, Timo Sirainen wrote:
> Which is probably a few seconds, so I don't see this as much of a problem.
Ok, thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
ile1 to filer2) :
. Maildir in source : message tagged as New
. initial dsync
. user read the message in the source, message is now tagges as Read
. switch user to Maildir in destination
. kick user
. user reconnects and sees, in destination, the message he just read tagged
as New as long
On Wed, Apr 10, 2013 at 06:28:47PM +0200, Thomas Hummel wrote:
> So basically, it works as with conventional data moved with rsync in 2 phases
> (initial copy and sync of the (hopefully small) reminder to minimize
> «downtime»
> or incomplete state except that dsync is used to have
ge, right ?
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
pace settings and a dsync copy but I
cannot figure out how exactly.
What do you think ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
On Tue, Feb 26, 2013 at 02:04:07PM +0100, Thomas Hummel wrote:
> Besides, why would [...]
I think I get it, I might have been confused by the term "unlimited" in "the
only useful values are 0 (unlimited)" in the sample config file comment.
As I see it now :
- service_co
On Tue, Feb 26, 2013 at 11:08:18AM +0100, Thomas Hummel wrote:
> -> Does it make sense
Sorry, I was unclear. Let me rephrase :
from my understanding of the doc, process_min_avail seems to have a slightly
different behavior depending of if service_count is 1 or 0. Is it correct ?
The doc
000
default_client_limit,
right ?
-> is 1024M too high for vsz_limit and may expose to memory leaks ?
In other words, what limit should be raised to scale a lot a connections ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
mailboxes on NFS, I'd want 2 LMTP servers :
on which servers should those LMTP servers run ? On the director servers (which
would mean director should have the mailboxes mounted) ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
posit on the mailbox since, by definition,
the user has not done any request yet ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
ess I still have the
option to use the litescape plugin.
Besides, how would one process to do the conversion ? Using dsync ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
lock users who
. read or write to much / slice of time
. login/logout too often / slice of time
Note : I'm still using the 1.2.x branch.
Thanks
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
re comparing it to the LDAP crypted
password
or does dovecot try to simply bind to LDAP with the clear password ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Groupe Exploitation et Infrastructure
t; }
Oh yes, thanks.
Also, is tcpwrap compatible with service_count = 0 in imap-login {} service ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
ppers) but I get :
doveot: imap-login: Error: connect(tcpwrap) failed: No such file or directory
Besides, my understanding is that with dovecot linked to libwrap, I can
avoid spawning imap-login through inetd. Is that correct ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle
ion level) some
sources ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
or multiple address on the same "address" line.
2. besides I don't know if it's possible to listen on the same address to
several ports for the same "subservice" : for instance could I have an
inet_listener imap {} listen on zz:zz:zz:zz:port1 and zz:zz:zz:zz:port2 ?
On Thu, Jan 13, 2011 at 02:51:50PM +0100, Thomas Hummel wrote:
> On Thu, Jan 13, 2011 at 02:07:47PM +0100, Pascal Volk wrote:
>
> > when you are logged in to your box as user 'thomas', then thomas is the
> > logged in system user.
>
> Ah sorry, I had int
On Thu, Jan 13, 2011 at 02:07:47PM +0100, Pascal Volk wrote:
> when you are logged in to your box as user 'thomas', then thomas is the
> logged in system user.
Ah sorry, I had interpreted "logged in" as "logged in through IMAP"...
--
Thomas Hummel | I
Hello,
Silly question : what user is the "logged in system user" doveadm-expunge's man
page refers to when describing its first form of use ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Tue, Jan 11, 2011 at 03:46:24PM +0100, Thomas Hummel wrote:
> Also, I'm not sure I correctly understand what "Use the default namespace for
> saving subscriptions." option is. Is it just the location of the subscription
> file for mailboxes subscribed in the shared na
dden = no
# list = no
list = children
}
# pwd
/courriel/boites/doveimap
# cat dovecot-acl-list
1294755382 Common
# cat .Common/dovecot-acl
user=dovepop akxeilprwts
# cat .Private/dovecot-acl
cat: .Private/dovecot-acl: No such file or directory
Also, I'm not sure I correctly understand what "Use the default namespace for
saving subscriptions." option is. Is it just the location of the subscription
file for mailboxes subscribed in the shared namespace ? Where would it be if
set to "yes" ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Thu, Dec 23, 2010 at 02:19:22PM +0100, Thomas Hummel wrote:
> 20101223-135556-38406.in:5 select "AA/BB/CC/foobar"
I forgot to mention that some of those mailboxes were not touched at all by the
user during the period where dovecot was using "/". And now that dovecot
On Thu, Dec 23, 2010 at 02:19:22PM +0100, Thomas Hummel wrote:
Also, if we change the separator, do we have to change it on the subscription
file ? Could it be the source of my problem ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Wed, Dec 22, 2010 at 10:03:00PM +0100, Thomas Hummel wrote:
So basically, I went from
1.2 with
#separator =
-> can you confirm the the client should use "." then ?
to
2.8.0 with
namespace {
type = private
separator = "/"
and now back to
On Wed, Dec 22, 2010 at 09:52:51PM +0100, Thomas Hummel wrote:
> in 1.2.12, I left the separator= commented out, so I had the default.
> in 2.0.8 I set separator = / myself.
>
> Still I can't see why it would lead TB to use a dot in a previously existing
> mailbox name.
Coul
ad TB to use a dot in a previously existing
mailbox name.
The separator is just the separator used by the client in the mailboxnames when
it sends commands to the server, right ?
> Do you have any shared/public namespaces?
Yes, one shared namespace, with separator = / too.
--
Thomas Hummel
to do.
> http://wiki2.dovecot.org/Debugging/Rawlog
Should I restart dovecot or just a SIGHUP ?
My other questions remains :
. can we revert from 2.0.8 to 1.2.12 (just in case) ?
. is there, between 1.2.12 and 2.0.8 some noticeable behavior regarding
mailbox name encoding ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Wed, Dec 22, 2010 at 05:38:06PM +0100, Thomas Hummel wrote:
> On Wed, Dec 22, 2010 at 05:28:07PM +0100, Thomas Hummel wrote:
>
> > hmmm, not so sure as some other user states the message (which I still
> > haven't
> > seen myself) talks about "." as the
On Wed, Dec 22, 2010 at 05:28:07PM +0100, Thomas Hummel wrote:
> hmmm, not so sure as some other user states the message (which I still haven't
> seen myself) talks about "." as the invalid character.
And finally one case correctly described :
A user couldn't move a m
On Wed, Dec 22, 2010 at 04:42:28PM +0100, Thomas Hummel wrote:
> However, some info seems to confirm that it involves moving a message from a
> mailbox to another where one of the mailbox has got non-ASCII character. A
> restart of TB seems to fix the problem indeed.
hmmm, not so sur
On Wed, Dec 22, 2010 at 04:13:20PM +0100, Thomas Hummel wrote:
> This is user input so I don't have any details nor sufficient element to
> consider it a real problem. A Thunderbird restart seems to solve the problem.
However, some info seems to confirm that it involves moving a mes
ng goes really wrong, is it safe to revert to 1.2.12 ? Maybe some
files has non-compatible format (dovecot-uidlist, indexes, ...) ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
n the "client_limit" setting ? For instance,
if a client (User Agent like Thunderbird) is configured to open 5 connections,
does that increase the client count by 5 for this limit ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
nyway
> better to set it too high than too low to avoid killing existing connections
> when vsz_limit is exceeded.
Just to be sure : what is that vsz_limit exactly to begin with ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
z_limit should be increased to avoid out of memory errors, especially if
you're using SSL/TLS."
Is there a way/rule of thumb to guess the appropriate value ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Fri, Nov 26, 2010 at 03:58:12PM +, Timo Sirainen wrote:
> No, you don't. dovecot-lda and Dovecot's LMTP server uses exactly the
> same code for writing mails.
Ok. Thanks. It would have been weird not to use such an advantage (early
indexing)
in a director environment.
-
On Fri, Nov 26, 2010 at 03:16:44PM +, Timo Sirainen wrote:
> On Fri, 2010-11-26 at 15:11 +0100, Thomas Hummel wrote:
> > Or is the LMTP server able to directly put the message in the final mailbox
> > ?
>
> Yes. LMTP server could be thought of as LDA.
But then you lo
ht interpret as "an SMTP server connects via LMTP to an LDA". Or is
the LMTP server able to directly put the message in the final mailbox ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
and dovecot-uidlist, wouldn't it need to be
proxied as well through director ? Is that the case ? If not, how can we use
deliver as LDA in a setup where n directors would proxy to m dovecot backends ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
are dovecot -n output and a gdb backtrace of the core file.
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
# 1.2.12: /usr/local/dovecot-1.2.12/etc/dovecot.conf
# OS: FreeBSD 7.2-STABLE amd64 nfs
syslog_facility: local0
protocols: imap imaps pop3 pop3s
hidden=yes, won't even see the default
namespace."
But if so, that would be weird that a bad client would be able to list
mailboxes and not a good client ?
Or said in another way, I don't understand why list=yes isn't enough in
anycase to list a namespace content.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Mon, Aug 09, 2010 at 10:56:21PM +0100, Timo Sirainen wrote:
> Because listing mailboxes always starts with the namespace that has no
> prefix. If you have only one namespace with prefix="" and list=no, you
> wouldn't
> have any listable mailboxes anywhere.
> For example if you have:
>
> name
On Mon, Aug 09, 2010 at 06:17:06PM +0100, Timo Sirainen wrote:
> On Mon, 2010-08-09 at 19:12 +0200, Thomas Hummel wrote:
> > > > > I've noticed that Thunderbird (3.0.3) seems to wrongly "repeat" the
> > > > > namespace
> > > >
On Wed, Jun 09, 2010 at 06:58:40PM +0200, Thomas Hummel wrote:
> On Tue, May 25, 2010 at 07:04:13PM +0100, Timo Sirainen wrote:
> > On Thu, 2010-04-22 at 15:31 +0200, Thomas Hummel wrote:
> >
> > > I've noticed that Thunderbird (3.0.3) seems to wrongly "repeat&
line 1: Unknown setting:
drop_priv_before_exec
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
t-2/
and I'm doing :
cd /usr/local/dovecot-2/bin
./doveconf -c -n /usr/local/dovecot/etc/dovecot.conf
...
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
gt; config file so I could get this fixed?
Here it is as an attachment.
> default_client_limit: This is mainly for non-imap/pop3 processes, since
> they have overridden it. 1 client = 1 connection, usually from another
> Dovecot process.
You mean internal dovecot processes other than mast
fault in the 10-master.conf file ?
>
> The defaults also have one example auth(-client) socket commented out
> for Postfix. You can create more of them if you want, but unless
> something actually uses them they're a bit pointless.
Ok, maybe I meant "the auth-userdb" unix listener is mandatory" : is it
?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
b process' ?
Same question : what is the auth-client socket used for ?
Finally, would it make sense to declare other auth listeners than the two
listed by default in the 10-master.conf file ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Fri, Jun 11, 2010 at 05:26:30PM +0100, Timo Sirainen wrote:
> > Correct ?
>
> Yes.
Thanks a lot Timo !
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
-> a setup where we don't want to support namespaces at all
Correct ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
t too". And since all namespaces are default
namespace children (since it's the one with an empty prefix), any
namespace with "list=yes" will be listed (and their content) by a list
"" * command, correct ?
Or, said in a simple way : list=yes == list "" * breaks name
f it doesn't support NAMESPACE or
doesn't have the namespace prefix somewhere in its conf.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
in the namespace only if we somehow know the name of the namespace and
specify it in the list command.
You seem to imply that a client who would do list "" * would list the
mailboxes of all namespaces configured with list=yes. I thought it would
list only the default namespace mailboxe
On Fri, Jun 11, 2010 at 02:48:57PM +0100, Timo Sirainen wrote:
> On to, 2010-06-10 at 18:18 +0200, Thomas Hummel wrote:
> > > See this thread:
> > > http://mailman2.u.washington.edu/pipermail/imap-protocol/2010-May/001076.html
> > >
> >
> > I&
On Wed, Jun 09, 2010 at 06:17:49PM +0100, Timo Sirainen wrote:
> On ke, 2010-06-09 at 18:58 +0200, Thomas Hummel wrote:
> > On Tue, May 25, 2010 at 07:04:13PM +0100, Timo Sirainen wrote:
> > > On Thu, 2010-04-22 at 15:31 +0200, Thomas Hummel wrote:
> > >
> >
On Tue, May 25, 2010 at 07:04:13PM +0100, Timo Sirainen wrote:
> On Thu, 2010-04-22 at 15:31 +0200, Thomas Hummel wrote:
>
> > I've noticed that Thunderbird (3.0.3) seems to wrongly "repeat" the
> > namespace
> > prefix when selecting the top level maildir
= no
> }
Thanks. I have to look further in what is allowed to go into those sections.
But in the case of disabling/enabling STARTTLS, this is indeed what I need.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
(hidden=, list=, "Allow Server to
override [the namesapce] settings") where TB knows about the namespace.
Any ideas ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
erent user agent of even the same user
? Can't we end up in a message tagged as "important" in a client being tagged as
"personnal" in another client ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
ng
for
another client (user2), will they ?
[Note : isn't it the same event with no shared mailboxes but with one user
using several clients ?]
But, if they're private, they have to be stored outside the filename : how
can we do that ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Fri, Apr 09, 2010 at 11:17:26AM +0200, Thomas Hummel wrote:
> > and the 'dovecot-shared' file doesn't help (besides, my understanding is
> > that
> > in 1.2x, it's for backward compatibility reason).
Which makes me wonder : does the dovecot-shared file s
On Wed, Mar 31, 2010 at 07:40:29PM +0200, Thomas Hummel wrote:
> Hello Timo,
>
> I'm running dovecot-1.2.11/Maildir and plan to migrate to single UID mailboxes
> some day, but for now, I've got "system" users and I'm testing permissions
> handling in or
quot;
* LIST (\HasNoChildren) "/" "INBOX"
* LIST (\Noselect \HasChildren) "/" "#shared"
19 OK List completed.
* LIST (\Noselect \HasChildren) "/" "#shared/doveimap"
20 OK List completed.
21 OK Lsub completed.
* LIST (\Noselect \HasCh
nderstanding is that
in 1.2x, it's for backward compatibility reason).
Any idea ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
the
"system_groups_user" extra-field ? As I see it now, system_groups_user gives
finer grain control since it's on a user basis.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Mon, Mar 22, 2010 at 03:40:23PM +0200, Timo Sirainen wrote:
> On Mon, 2010-03-22 at 14:34 +0100, Thomas Hummel wrote:
>
> > You lost me again : what's the use of the userdb " acl_groups" extra-field
> > then ?
>
> That only adds user to given ACL group
you deal with that
> > ?
>
> You don't need to put all users into doveshared group. You just need to
> set mail_extra_groups=doveshared, so only Dovecot processes have such
> extra access.
You lost me again : what's the use of the userdb " acl_groups" extra-f
give access to the mailboxes
by anything else than IMAP. If so, isn't it quite the same as 0777 with a
mail_location outside of user's reach (except through IMAP) ?
You maybe would you rely on filesystem acls to have a finer grain access
control ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
ny groups (similar in purpose to your
doveshared one) as needs to share a mailbox, if and only if I could somehow
restrict (politically I mean) the use of shared mailboxes to "privileged" users
(for instance a unit chief and his assistant, ...). Not really scalable I'm
afraid thou
t the shared-mailbox file format and try to hack it directly,
right ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
in the system user case ? (I
mean on an OS level, problems related to such loose permission are the same
and, if in 777, permissions are no longer a problem for dovecot) ?
Thanks.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Thu, Mar 18, 2010 at 05:50:44PM +0100, Thomas Hummel wrote:
> But isn't a client like Thunderbird (3.0.3) supposed to
>
> . show me the #shared (even if no one shares mailboxes to me) "folder"
> (because of list = yes )?
>
> . let me share, let'
the #shared (even if no one shares mailboxes to me) "folder"
(because of list = yes )?
. let me share, let's say the .sous.arbo mailbox ? If yes, how ? Or should I
first pre-set some acl files ???
For now, it says : "This is a personnal mail folder. It is not shared."
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
map-login)
...
. callback : a pointer in a lower level to a function in an higher level
. context : what links "objects" of those layers eachother
I was thinking that, if my understanding is correct, some overview from you of
such an architecture would be of great help for those
ith multiple alt subjects that should cover all the names that server
> may have. The client should support those kind of certificates, of
> course.
Thanks.
You're not talking about wildcard certificate, aren't you ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
e
than
one instance of dovecot (several dovecot with different config files).
The problem is that some clients may be configured with mail.my.domain, some
others with imap.my.domain, ...etc... Hence the need to have different
certificates with those different names as cn.
--
Thomas Hummel | I
On Wed, Oct 21, 2009 at 09:39:22AM -0700, Brandon Davidson wrote:
> As a contrasting data point, we run NFS + random redirects with almost no
> problems.
Thanks for your answer as well.
What mailbox format are you using ?
--
Thomas Hummel | Institut Pasteur
| Pôle inform
On Wed, Oct 21, 2009 at 04:59:50PM +0100, Guy wrote:
> Our current setup uses two NFS mounts accessed simultaneously by two
> servers.
[...]
Thanks for sharing your experience.
Are you using mbox, dbox or maildir ?
What % of IMAP and POP3 clients ?
--
Thomas Hummel | Institut P
On Tue, Oct 20, 2009 at 10:47:25AM +0200, Thomas Hummel wrote:
> > Actual mail content should be safe.
So you seem to say that indexes files would probably get "corrupted" but that
clients wouldn't notice it ?
I'm trying to figure out how to use imap-test test sc
ldn't
be corrupted I guess, so I would that be recoverable on the client ?
> The long term fix for this is
> http://www.dovecot.org/list/dovecot/2009-August/041983.html
Thanks.
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
rough 2
dovecot servers ? I don't see imaptest.c anymore on
http://dovecot.org/tools
but I see 2 nfs_test ? Is there some command line options I should use ?
6. when and why can a random redirect to multiple server can cause mailbox
corruption ? On flags only or on content as well ?
Thanks
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
On Fri, Jul 31, 2009 at 01:34:59PM -0400, Timo Sirainen wrote:
Thanks for your explanations and patience Timo!
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
ing like, in its cache, the client has a mapping like
slot 1 <-> UID u
slot 2 <-> UID v
...
and he sees
slot 1 <-> UID z ?
--
Thomas Hummel | Institut Pasteur
| Pôle informatique - systèmes et réseau
1 - 100 of 161 matches
Mail list logo
