You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@couchdb.apache.org by Martin Hewitt <ma...@thenoi.se> on 2012/10/03 13:35:02 UTC
eacces error
Hi all,
I've recently compiled CouchDB from source, on an EC2 instance, and I'm seeing a lot of eacces errors of the form:
[Wed, 03 Oct 2012 11:32:22 GMT] [error] [<0.19.0>] {error_report,<0.9.0>,
{<0.19.0>,std_error,
"File operation error: eacces. Target: ./mochiweb.beam. Function: get_file. Process: code_server."}}
Now I know this is a permissions issue, but I can't work out what permissions I need to change.
I've changed the owner on the data directories, ERLang directories any many more besides, but I'm still getting the error.
If anyone could point me in the right direction, I'd be hugely grateful.
Thanks,
Martin
Re: eacces error
Posted by Noah Slater <ns...@tumbolia.org>.
`dirname $0` in sh.
I am confused here. There are three things which are candidates for
permission errors.
The init.d script attempts to execute `couchdb` from the `couchdb` user's
shell.
The `couchdb` script itself.
The Erlang process.
When Bob mentioned the problem about not being in a guaranteed readable
dir, which one did he mean?
The `couchdb` script itself sometimes spits out files into the `pwd` so we
can't have it automatically switch dirs.
If it is the Erlang process itself, then that is another matter entirely.
Do we have to wait for OTP for that?
And what has the shell synatax Benoit quotes have to do with anything?
Sorry for my confusion.
On Sun, Oct 7, 2012 at 8:38 PM, Dave Cottlehuber <dc...@jsonified.com> wrote:
> On 7 October 2012 21:18, Benoit Chesneau <bc...@gmail.com> wrote:
> > On Sun, Oct 7, 2012 at 6:59 PM, Noah Slater <ns...@tumbolia.org>
> wrote:
> >> Right, but realistically, it may be a number of months before we have an
> >> OTP release. So a simple patch might be nice for now. Dave's proposal
> looks
> >> sane to me.
> >>
> >> What is a Dix? And how does an OTP release fix the problem? Anyone?
> Excuse
> >> me for my ignorance.
> >
> > Not specific to otp but maybe that change could do the trick:
> >
> > https://github.com/refuge/rcouch/blob/master/rel/files/rcouch#L6
>
> Thanks Benoit,
>
> I put it into a simple shell script & it only returns the command-name
> for me. How do you extract the runtime base dir using that syntax? I'm
> assuming this will be embarassingly obvious.
>
> A+
> Dave
>
--
NS
Re: eacces error
Posted by Dave Cottlehuber <dc...@jsonified.com>.
On 7 October 2012 21:18, Benoit Chesneau <bc...@gmail.com> wrote:
> On Sun, Oct 7, 2012 at 6:59 PM, Noah Slater <ns...@tumbolia.org> wrote:
>> Right, but realistically, it may be a number of months before we have an
>> OTP release. So a simple patch might be nice for now. Dave's proposal looks
>> sane to me.
>>
>> What is a Dix? And how does an OTP release fix the problem? Anyone? Excuse
>> me for my ignorance.
>
> Not specific to otp but maybe that change could do the trick:
>
> https://github.com/refuge/rcouch/blob/master/rel/files/rcouch#L6
Thanks Benoit,
I put it into a simple shell script & it only returns the command-name
for me. How do you extract the runtime base dir using that syntax? I'm
assuming this will be embarassingly obvious.
A+
Dave
Re: eacces error
Posted by Benoit Chesneau <bc...@gmail.com>.
On Sun, Oct 7, 2012 at 6:59 PM, Noah Slater <ns...@tumbolia.org> wrote:
> Right, but realistically, it may be a number of months before we have an
> OTP release. So a simple patch might be nice for now. Dave's proposal looks
> sane to me.
>
> What is a Dix? And how does an OTP release fix the problem? Anyone? Excuse
> me for my ignorance.
Not specific to otp but maybe that change could do the trick:
https://github.com/refuge/rcouch/blob/master/rel/files/rcouch#L6
>
> On Thu, Oct 4, 2012 at 5:46 PM, Robert Newson <ro...@gmail.com>wrote:
>
>> Moving to a proper OTP release *is* the fix. There's a Dix on the two
>> ways erlang starts. One is called "interactive" and is how couch and,
>> I think, bigcouch are started. That mode is the one that includes your
>> current working dir. The other mode, the name escapes me, does not.
>>
>> Sent from the ocean floor
>>
>> On 4 Oct 2012, at 15:13, Dave Cottlehuber <dc...@jsonified.com> wrote:
>>
>> > On 4 October 2012 12:58, Noah Slater <ns...@tumbolia.org> wrote:
>> >> Should we patch this before the move to an OTP system? Does anybody
>> have a
>> >> patch?
>> >
>> > I'll take a look, something like chdir to /usr/local/var/run/couchdb
>> > or similar might be an approach. I think on Windows at least it
>> > matters where CouchDB is running from, but it uses a different method
>> > of starting up vs *nix so that should be OK.
>> >
>> > @Martin does this patch fix it? Just ensure your /usr/* paths match up.
>> >
>> > --- /usr/local/bin/couchdb 2012-10-04 16:06:16.000000000 +0200
>> > +++ /tmp/couchdb 2012-10-04 16:06:02.000000000 +0200
>> > @@ -27,6 +27,7 @@
>> > KILL=false
>> > LOCAL_CONFIG_DIR=/usr/local/etc/couchdb/local.d
>> > LOCAL_CONFIG_FILE=/usr/local/etc/couchdb/local.ini
>> > +LOCAL_RUN_DIR=/usr/local/var/run/couchdb/
>> > PID_FILE=/usr/local/var/run/couchdb/couchdb.pid
>> > RECURSED=false
>> > RESET_CONFIG=true
>> > @@ -220,13 +221,16 @@
>> > check_environment
>> > fi
>> > interactive_option="+Bd -noinput"
>> > + working_dir=$PWD
>> > if test "$INTERACTIVE" = "true"; then
>> > interactive_option=""
>> > + working_dir=$LOCAL_RUN_DIR
>> > fi
>> > if test "$BACKGROUND" = "true"; then
>> > touch $PID_FILE
>> > interactive_option="+Bd -noinput"
>> > fi
>> > + cd $working_dir
>> > command="/usr/local/opt/erlang/bin/erl $interactive_option
>> > $ERL_START_OPTIONS \
>> > -env ERL_LIBS
>> > /usr/local/Cellar/couchdb/HEAD/lib/couchdb/erlang/lib -couch_ini
>> > $start_arguments -s couch"
>> > if test "$BACKGROUND" = "true" -a "$RECURSED" = "false"; then
>> >
>> > I've got a patch against master if this works.
>> >
>> > A+
>> > Dave
>>
>
>
>
> --
> NS
Re: eacces error
Posted by Robert Newson <ro...@gmail.com>.
Err, "doc". I just mean that there's two ways to start up, and they
vary by the method they search for modules.
Simplest fix is a 'cd' to any dir readable by the couchdb user before 'erl'.
Sent from the ocean floor
On 7 Oct 2012, at 18:00, Noah Slater <ns...@tumbolia.org> wrote:
> Right, but realistically, it may be a number of months before we have an
> OTP release. So a simple patch might be nice for now. Dave's proposal looks
> sane to me.
>
> What is a Dix? And how does an OTP release fix the problem? Anyone? Excuse
> me for my ignorance.
>
> On Thu, Oct 4, 2012 at 5:46 PM, Robert Newson <ro...@gmail.com>wrote:
>
>> Moving to a proper OTP release *is* the fix. There's a Dix on the two
>> ways erlang starts. One is called "interactive" and is how couch and,
>> I think, bigcouch are started. That mode is the one that includes your
>> current working dir. The other mode, the name escapes me, does not.
>>
>> Sent from the ocean floor
>>
>> On 4 Oct 2012, at 15:13, Dave Cottlehuber <dc...@jsonified.com> wrote:
>>
>>> On 4 October 2012 12:58, Noah Slater <ns...@tumbolia.org> wrote:
>>>> Should we patch this before the move to an OTP system? Does anybody
>> have a
>>>> patch?
>>>
>>> I'll take a look, something like chdir to /usr/local/var/run/couchdb
>>> or similar might be an approach. I think on Windows at least it
>>> matters where CouchDB is running from, but it uses a different method
>>> of starting up vs *nix so that should be OK.
>>>
>>> @Martin does this patch fix it? Just ensure your /usr/* paths match up.
>>>
>>> --- /usr/local/bin/couchdb 2012-10-04 16:06:16.000000000 +0200
>>> +++ /tmp/couchdb 2012-10-04 16:06:02.000000000 +0200
>>> @@ -27,6 +27,7 @@
>>> KILL=false
>>> LOCAL_CONFIG_DIR=/usr/local/etc/couchdb/local.d
>>> LOCAL_CONFIG_FILE=/usr/local/etc/couchdb/local.ini
>>> +LOCAL_RUN_DIR=/usr/local/var/run/couchdb/
>>> PID_FILE=/usr/local/var/run/couchdb/couchdb.pid
>>> RECURSED=false
>>> RESET_CONFIG=true
>>> @@ -220,13 +221,16 @@
>>> check_environment
>>> fi
>>> interactive_option="+Bd -noinput"
>>> + working_dir=$PWD
>>> if test "$INTERACTIVE" = "true"; then
>>> interactive_option=""
>>> + working_dir=$LOCAL_RUN_DIR
>>> fi
>>> if test "$BACKGROUND" = "true"; then
>>> touch $PID_FILE
>>> interactive_option="+Bd -noinput"
>>> fi
>>> + cd $working_dir
>>> command="/usr/local/opt/erlang/bin/erl $interactive_option
>>> $ERL_START_OPTIONS \
>>> -env ERL_LIBS
>>> /usr/local/Cellar/couchdb/HEAD/lib/couchdb/erlang/lib -couch_ini
>>> $start_arguments -s couch"
>>> if test "$BACKGROUND" = "true" -a "$RECURSED" = "false"; then
>>>
>>> I've got a patch against master if this works.
>>>
>>> A+
>>> Dave
>
>
>
> --
> NS
Re: eacces error
Posted by Noah Slater <ns...@tumbolia.org>.
Right, but realistically, it may be a number of months before we have an
OTP release. So a simple patch might be nice for now. Dave's proposal looks
sane to me.
What is a Dix? And how does an OTP release fix the problem? Anyone? Excuse
me for my ignorance.
On Thu, Oct 4, 2012 at 5:46 PM, Robert Newson <ro...@gmail.com>wrote:
> Moving to a proper OTP release *is* the fix. There's a Dix on the two
> ways erlang starts. One is called "interactive" and is how couch and,
> I think, bigcouch are started. That mode is the one that includes your
> current working dir. The other mode, the name escapes me, does not.
>
> Sent from the ocean floor
>
> On 4 Oct 2012, at 15:13, Dave Cottlehuber <dc...@jsonified.com> wrote:
>
> > On 4 October 2012 12:58, Noah Slater <ns...@tumbolia.org> wrote:
> >> Should we patch this before the move to an OTP system? Does anybody
> have a
> >> patch?
> >
> > I'll take a look, something like chdir to /usr/local/var/run/couchdb
> > or similar might be an approach. I think on Windows at least it
> > matters where CouchDB is running from, but it uses a different method
> > of starting up vs *nix so that should be OK.
> >
> > @Martin does this patch fix it? Just ensure your /usr/* paths match up.
> >
> > --- /usr/local/bin/couchdb 2012-10-04 16:06:16.000000000 +0200
> > +++ /tmp/couchdb 2012-10-04 16:06:02.000000000 +0200
> > @@ -27,6 +27,7 @@
> > KILL=false
> > LOCAL_CONFIG_DIR=/usr/local/etc/couchdb/local.d
> > LOCAL_CONFIG_FILE=/usr/local/etc/couchdb/local.ini
> > +LOCAL_RUN_DIR=/usr/local/var/run/couchdb/
> > PID_FILE=/usr/local/var/run/couchdb/couchdb.pid
> > RECURSED=false
> > RESET_CONFIG=true
> > @@ -220,13 +221,16 @@
> > check_environment
> > fi
> > interactive_option="+Bd -noinput"
> > + working_dir=$PWD
> > if test "$INTERACTIVE" = "true"; then
> > interactive_option=""
> > + working_dir=$LOCAL_RUN_DIR
> > fi
> > if test "$BACKGROUND" = "true"; then
> > touch $PID_FILE
> > interactive_option="+Bd -noinput"
> > fi
> > + cd $working_dir
> > command="/usr/local/opt/erlang/bin/erl $interactive_option
> > $ERL_START_OPTIONS \
> > -env ERL_LIBS
> > /usr/local/Cellar/couchdb/HEAD/lib/couchdb/erlang/lib -couch_ini
> > $start_arguments -s couch"
> > if test "$BACKGROUND" = "true" -a "$RECURSED" = "false"; then
> >
> > I've got a patch against master if this works.
> >
> > A+
> > Dave
>
--
NS
Re: eacces error
Posted by Robert Newson <ro...@gmail.com>.
Moving to a proper OTP release *is* the fix. There's a Dix on the two
ways erlang starts. One is called "interactive" and is how couch and,
I think, bigcouch are started. That mode is the one that includes your
current working dir. The other mode, the name escapes me, does not.
Sent from the ocean floor
On 4 Oct 2012, at 15:13, Dave Cottlehuber <dc...@jsonified.com> wrote:
> On 4 October 2012 12:58, Noah Slater <ns...@tumbolia.org> wrote:
>> Should we patch this before the move to an OTP system? Does anybody have a
>> patch?
>
> I'll take a look, something like chdir to /usr/local/var/run/couchdb
> or similar might be an approach. I think on Windows at least it
> matters where CouchDB is running from, but it uses a different method
> of starting up vs *nix so that should be OK.
>
> @Martin does this patch fix it? Just ensure your /usr/* paths match up.
>
> --- /usr/local/bin/couchdb 2012-10-04 16:06:16.000000000 +0200
> +++ /tmp/couchdb 2012-10-04 16:06:02.000000000 +0200
> @@ -27,6 +27,7 @@
> KILL=false
> LOCAL_CONFIG_DIR=/usr/local/etc/couchdb/local.d
> LOCAL_CONFIG_FILE=/usr/local/etc/couchdb/local.ini
> +LOCAL_RUN_DIR=/usr/local/var/run/couchdb/
> PID_FILE=/usr/local/var/run/couchdb/couchdb.pid
> RECURSED=false
> RESET_CONFIG=true
> @@ -220,13 +221,16 @@
> check_environment
> fi
> interactive_option="+Bd -noinput"
> + working_dir=$PWD
> if test "$INTERACTIVE" = "true"; then
> interactive_option=""
> + working_dir=$LOCAL_RUN_DIR
> fi
> if test "$BACKGROUND" = "true"; then
> touch $PID_FILE
> interactive_option="+Bd -noinput"
> fi
> + cd $working_dir
> command="/usr/local/opt/erlang/bin/erl $interactive_option
> $ERL_START_OPTIONS \
> -env ERL_LIBS
> /usr/local/Cellar/couchdb/HEAD/lib/couchdb/erlang/lib -couch_ini
> $start_arguments -s couch"
> if test "$BACKGROUND" = "true" -a "$RECURSED" = "false"; then
>
> I've got a patch against master if this works.
>
> A+
> Dave
Re: eacces error
Posted by Dave Cottlehuber <dc...@jsonified.com>.
On 4 October 2012 12:58, Noah Slater <ns...@tumbolia.org> wrote:
> Should we patch this before the move to an OTP system? Does anybody have a
> patch?
I'll take a look, something like chdir to /usr/local/var/run/couchdb
or similar might be an approach. I think on Windows at least it
matters where CouchDB is running from, but it uses a different method
of starting up vs *nix so that should be OK.
@Martin does this patch fix it? Just ensure your /usr/* paths match up.
--- /usr/local/bin/couchdb 2012-10-04 16:06:16.000000000 +0200
+++ /tmp/couchdb 2012-10-04 16:06:02.000000000 +0200
@@ -27,6 +27,7 @@
KILL=false
LOCAL_CONFIG_DIR=/usr/local/etc/couchdb/local.d
LOCAL_CONFIG_FILE=/usr/local/etc/couchdb/local.ini
+LOCAL_RUN_DIR=/usr/local/var/run/couchdb/
PID_FILE=/usr/local/var/run/couchdb/couchdb.pid
RECURSED=false
RESET_CONFIG=true
@@ -220,13 +221,16 @@
check_environment
fi
interactive_option="+Bd -noinput"
+ working_dir=$PWD
if test "$INTERACTIVE" = "true"; then
interactive_option=""
+ working_dir=$LOCAL_RUN_DIR
fi
if test "$BACKGROUND" = "true"; then
touch $PID_FILE
interactive_option="+Bd -noinput"
fi
+ cd $working_dir
command="/usr/local/opt/erlang/bin/erl $interactive_option
$ERL_START_OPTIONS \
-env ERL_LIBS
/usr/local/Cellar/couchdb/HEAD/lib/couchdb/erlang/lib -couch_ini
$start_arguments -s couch"
if test "$BACKGROUND" = "true" -a "$RECURSED" = "false"; then
I've got a patch against master if this works.
A+
Dave
Re: eacces error
Posted by Noah Slater <ns...@tumbolia.org>.
Should we patch this before the move to an OTP system? Does anybody have a
patch?
On Wed, Oct 3, 2012 at 6:21 PM, Robert Newson <ro...@gmail.com>wrote:
> Yup. This is the silly thing where, in interactive mode, erlang will
> such current working for for modules. Since we don't "cd" in the
> script to a guaranteed readable dir, spurious errors abound.
>
> A proper OTP launch would not do this as all paths are defined up front.
>
> Sent from the pool side with a glass of wine in hand.
>
> On 3 Oct 2012, at 17:00, Adam Kocoloski <ko...@apache.org> wrote:
>
> > On Oct 3, 2012, at 10:54 AM, Dave Cottlehuber <dc...@jsonified.com> wrote:
> >
> >> On 3 October 2012 16:49, Adam Kocoloski <ko...@apache.org> wrote:
> >>> Yep, that matches up with my experience.
> >>>
> >>> Adam
> >>>
> >>> On Oct 3, 2012, at 10:47 AM, Martin Hewitt <ma...@thenoi.se> wrote:
> >>>
> >>>> This one's further complicated by the fact that, after spitting out
> errors at start, CouchDB appears to operate normally with no further
> errors. Generating a view produces another error, but the view is generated
> correctly, and further view updates do not produce errors.
> >>>>
> >>>> Very curious behaviour.
> >>>>
> >>>> Martin
> >>
> >> Aaah this vaguely comes back to me now, hopefully enough to trigger
> >> somebody else's memory with something correct.
> >>
> >> I think it's due to the search path for the erlang code loader
> >> including directories it can't check/view. So if you can do a dtrace
> >> or truss or whatever , this should turn up pretty quickly.
> >>
> >> A+
> >> Dave
> >
> > Oh, that makes a ton of sense. Thanks for the tip!
>
--
NS
Re: eacces error
Posted by Robert Newson <ro...@gmail.com>.
Yup. This is the silly thing where, in interactive mode, erlang will
such current working for for modules. Since we don't "cd" in the
script to a guaranteed readable dir, spurious errors abound.
A proper OTP launch would not do this as all paths are defined up front.
Sent from the pool side with a glass of wine in hand.
On 3 Oct 2012, at 17:00, Adam Kocoloski <ko...@apache.org> wrote:
> On Oct 3, 2012, at 10:54 AM, Dave Cottlehuber <dc...@jsonified.com> wrote:
>
>> On 3 October 2012 16:49, Adam Kocoloski <ko...@apache.org> wrote:
>>> Yep, that matches up with my experience.
>>>
>>> Adam
>>>
>>> On Oct 3, 2012, at 10:47 AM, Martin Hewitt <ma...@thenoi.se> wrote:
>>>
>>>> This one's further complicated by the fact that, after spitting out errors at start, CouchDB appears to operate normally with no further errors. Generating a view produces another error, but the view is generated correctly, and further view updates do not produce errors.
>>>>
>>>> Very curious behaviour.
>>>>
>>>> Martin
>>
>> Aaah this vaguely comes back to me now, hopefully enough to trigger
>> somebody else's memory with something correct.
>>
>> I think it's due to the search path for the erlang code loader
>> including directories it can't check/view. So if you can do a dtrace
>> or truss or whatever , this should turn up pretty quickly.
>>
>> A+
>> Dave
>
> Oh, that makes a ton of sense. Thanks for the tip!
Re: eacces error
Posted by Adam Kocoloski <ko...@apache.org>.
On Oct 3, 2012, at 10:54 AM, Dave Cottlehuber <dc...@jsonified.com> wrote:
> On 3 October 2012 16:49, Adam Kocoloski <ko...@apache.org> wrote:
>> Yep, that matches up with my experience.
>>
>> Adam
>>
>> On Oct 3, 2012, at 10:47 AM, Martin Hewitt <ma...@thenoi.se> wrote:
>>
>>> This one's further complicated by the fact that, after spitting out errors at start, CouchDB appears to operate normally with no further errors. Generating a view produces another error, but the view is generated correctly, and further view updates do not produce errors.
>>>
>>> Very curious behaviour.
>>>
>>> Martin
>
> Aaah this vaguely comes back to me now, hopefully enough to trigger
> somebody else's memory with something correct.
>
> I think it's due to the search path for the erlang code loader
> including directories it can't check/view. So if you can do a dtrace
> or truss or whatever , this should turn up pretty quickly.
>
> A+
> Dave
Oh, that makes a ton of sense. Thanks for the tip!
Re: eacces error
Posted by Dave Cottlehuber <dc...@jsonified.com>.
On 3 October 2012 16:49, Adam Kocoloski <ko...@apache.org> wrote:
> Yep, that matches up with my experience.
>
> Adam
>
> On Oct 3, 2012, at 10:47 AM, Martin Hewitt <ma...@thenoi.se> wrote:
>
>> This one's further complicated by the fact that, after spitting out errors at start, CouchDB appears to operate normally with no further errors. Generating a view produces another error, but the view is generated correctly, and further view updates do not produce errors.
>>
>> Very curious behaviour.
>>
>> Martin
Aaah this vaguely comes back to me now, hopefully enough to trigger
somebody else's memory with something correct.
I think it's due to the search path for the erlang code loader
including directories it can't check/view. So if you can do a dtrace
or truss or whatever , this should turn up pretty quickly.
A+
Dave
Re: eacces error
Posted by Adam Kocoloski <ko...@apache.org>.
Yep, that matches up with my experience.
Adam
On Oct 3, 2012, at 10:47 AM, Martin Hewitt <ma...@thenoi.se> wrote:
> This one's further complicated by the fact that, after spitting out errors at start, CouchDB appears to operate normally with no further errors. Generating a view produces another error, but the view is generated correctly, and further view updates do not produce errors.
>
> Very curious behaviour.
>
> Martin
>
>
> On Wednesday, 3 October 2012 at 15:44, Adam Kocoloski wrote:
>
>> This one in particular has always been a bit of a mystery to me. Sometimes it seems to be completely spurious.
>>
>> Adam
>>
>> On Oct 3, 2012, at 8:08 AM, Martin Hewitt <martin@thenoi.se (mailto:martin@thenoi.se)> wrote:
>>
>>> Hi Dave,
>>>
>>> Thanks for these links, no luck on my current install, but I may scrub down and start again, as I can't really map what's going on, so probably won't be able to work out what makes a difference!
>>>
>>> Martin
>>>
>>>
>>> On Wednesday, 3 October 2012 at 13:01, Dave Cottlehuber wrote:
>>>
>>>> On 3 October 2012 13:35, Martin Hewitt <martin@thenoi.se (mailto:martin@thenoi.se)> wrote:
>>>>> Hi all,
>>>>>
>>>>> I've recently compiled CouchDB from source, on an EC2 instance, and I'm seeing a lot of eacces errors of the form:
>>>>>
>>>>> [Wed, 03 Oct 2012 11:32:22 GMT] [error] [<0.19.0>] {error_report,<0.9.0>,
>>>>> {<0.19.0>,std_error,
>>>>> "File operation error: eacces. Target: ./mochiweb.beam (http://web.beam). Function: get_file. Process: code_server."}}
>>>>>
>>>>>
>>>>> Now I know this is a permissions issue, but I can't work out what permissions I need to change.
>>>>>
>>>>> I've changed the owner on the data directories, ERLang directories any many more besides, but I'm still getting the error.
>>>>>
>>>>> If anyone could point me in the right direction, I'd be hugely grateful.
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Martin
>>>>
>>>> Most of the info is in here:
>>>> https://github.com/apache/couchdb/blob/master/INSTALL.Unix#L175-187
>>>> and some on the wiki
>>>> http://wiki.apache.org/couchdb/Installing_on_Ubuntu or similar ones
>>>> for other linux variants may help.
>>>>
>>>> su to the couch user, and then run couchdb interactively `couchdb -i`
>>>> which sometimes spits out a moment of clarity.
>>>>
>>>> IIRC there was a recent-ish example on the user@ mailing list but I
>>>> unfortunately don't recall the solution!
>>>>
>>>> A+
>>>> Dave
>>>>
>>>
>>>
>>
>>
>>
>
>
Re: eacces error
Posted by Martin Hewitt <ma...@thenoi.se>.
This one's further complicated by the fact that, after spitting out errors at start, CouchDB appears to operate normally with no further errors. Generating a view produces another error, but the view is generated correctly, and further view updates do not produce errors.
Very curious behaviour.
Martin
On Wednesday, 3 October 2012 at 15:44, Adam Kocoloski wrote:
> This one in particular has always been a bit of a mystery to me. Sometimes it seems to be completely spurious.
>
> Adam
>
> On Oct 3, 2012, at 8:08 AM, Martin Hewitt <martin@thenoi.se (mailto:martin@thenoi.se)> wrote:
>
> > Hi Dave,
> >
> > Thanks for these links, no luck on my current install, but I may scrub down and start again, as I can't really map what's going on, so probably won't be able to work out what makes a difference!
> >
> > Martin
> >
> >
> > On Wednesday, 3 October 2012 at 13:01, Dave Cottlehuber wrote:
> >
> > > On 3 October 2012 13:35, Martin Hewitt <martin@thenoi.se (mailto:martin@thenoi.se)> wrote:
> > > > Hi all,
> > > >
> > > > I've recently compiled CouchDB from source, on an EC2 instance, and I'm seeing a lot of eacces errors of the form:
> > > >
> > > > [Wed, 03 Oct 2012 11:32:22 GMT] [error] [<0.19.0>] {error_report,<0.9.0>,
> > > > {<0.19.0>,std_error,
> > > > "File operation error: eacces. Target: ./mochiweb.beam (http://web.beam). Function: get_file. Process: code_server."}}
> > > >
> > > >
> > > > Now I know this is a permissions issue, but I can't work out what permissions I need to change.
> > > >
> > > > I've changed the owner on the data directories, ERLang directories any many more besides, but I'm still getting the error.
> > > >
> > > > If anyone could point me in the right direction, I'd be hugely grateful.
> > > >
> > > > Thanks,
> > > >
> > > > Martin
> > >
> > > Most of the info is in here:
> > > https://github.com/apache/couchdb/blob/master/INSTALL.Unix#L175-187
> > > and some on the wiki
> > > http://wiki.apache.org/couchdb/Installing_on_Ubuntu or similar ones
> > > for other linux variants may help.
> > >
> > > su to the couch user, and then run couchdb interactively `couchdb -i`
> > > which sometimes spits out a moment of clarity.
> > >
> > > IIRC there was a recent-ish example on the user@ mailing list but I
> > > unfortunately don't recall the solution!
> > >
> > > A+
> > > Dave
> > >
> >
> >
>
>
>
Re: eacces error
Posted by Adam Kocoloski <ko...@apache.org>.
This one in particular has always been a bit of a mystery to me. Sometimes it seems to be completely spurious.
Adam
On Oct 3, 2012, at 8:08 AM, Martin Hewitt <ma...@thenoi.se> wrote:
> Hi Dave,
>
> Thanks for these links, no luck on my current install, but I may scrub down and start again, as I can't really map what's going on, so probably won't be able to work out what makes a difference!
>
> Martin
>
>
> On Wednesday, 3 October 2012 at 13:01, Dave Cottlehuber wrote:
>
>> On 3 October 2012 13:35, Martin Hewitt <martin@thenoi.se (mailto:martin@thenoi.se)> wrote:
>>> Hi all,
>>>
>>> I've recently compiled CouchDB from source, on an EC2 instance, and I'm seeing a lot of eacces errors of the form:
>>>
>>> [Wed, 03 Oct 2012 11:32:22 GMT] [error] [<0.19.0>] {error_report,<0.9.0>,
>>> {<0.19.0>,std_error,
>>> "File operation error: eacces. Target: ./mochiweb.beam (http://web.beam). Function: get_file. Process: code_server."}}
>>>
>>>
>>> Now I know this is a permissions issue, but I can't work out what permissions I need to change.
>>>
>>> I've changed the owner on the data directories, ERLang directories any many more besides, but I'm still getting the error.
>>>
>>> If anyone could point me in the right direction, I'd be hugely grateful.
>>>
>>> Thanks,
>>>
>>> Martin
>>
>> Most of the info is in here:
>> https://github.com/apache/couchdb/blob/master/INSTALL.Unix#L175-187
>> and some on the wiki
>> http://wiki.apache.org/couchdb/Installing_on_Ubuntu or similar ones
>> for other linux variants may help.
>>
>> su to the couch user, and then run couchdb interactively `couchdb -i`
>> which sometimes spits out a moment of clarity.
>>
>> IIRC there was a recent-ish example on the user@ mailing list but I
>> unfortunately don't recall the solution!
>>
>> A+
>> Dave
>>
>>
>
>
Re: eacces error
Posted by Martin Hewitt <ma...@thenoi.se>.
Hi Dave,
Thanks for these links, no luck on my current install, but I may scrub down and start again, as I can't really map what's going on, so probably won't be able to work out what makes a difference!
Martin
On Wednesday, 3 October 2012 at 13:01, Dave Cottlehuber wrote:
> On 3 October 2012 13:35, Martin Hewitt <martin@thenoi.se (mailto:martin@thenoi.se)> wrote:
> > Hi all,
> >
> > I've recently compiled CouchDB from source, on an EC2 instance, and I'm seeing a lot of eacces errors of the form:
> >
> > [Wed, 03 Oct 2012 11:32:22 GMT] [error] [<0.19.0>] {error_report,<0.9.0>,
> > {<0.19.0>,std_error,
> > "File operation error: eacces. Target: ./mochiweb.beam (http://web.beam). Function: get_file. Process: code_server."}}
> >
> >
> > Now I know this is a permissions issue, but I can't work out what permissions I need to change.
> >
> > I've changed the owner on the data directories, ERLang directories any many more besides, but I'm still getting the error.
> >
> > If anyone could point me in the right direction, I'd be hugely grateful.
> >
> > Thanks,
> >
> > Martin
>
> Most of the info is in here:
> https://github.com/apache/couchdb/blob/master/INSTALL.Unix#L175-187
> and some on the wiki
> http://wiki.apache.org/couchdb/Installing_on_Ubuntu or similar ones
> for other linux variants may help.
>
> su to the couch user, and then run couchdb interactively `couchdb -i`
> which sometimes spits out a moment of clarity.
>
> IIRC there was a recent-ish example on the user@ mailing list but I
> unfortunately don't recall the solution!
>
> A+
> Dave
>
>
Re: eacces error
Posted by Dave Cottlehuber <dc...@jsonified.com>.
On 3 October 2012 13:35, Martin Hewitt <ma...@thenoi.se> wrote:
> Hi all,
>
> I've recently compiled CouchDB from source, on an EC2 instance, and I'm seeing a lot of eacces errors of the form:
>
> [Wed, 03 Oct 2012 11:32:22 GMT] [error] [<0.19.0>] {error_report,<0.9.0>,
> {<0.19.0>,std_error,
> "File operation error: eacces. Target: ./mochiweb.beam. Function: get_file. Process: code_server."}}
>
>
> Now I know this is a permissions issue, but I can't work out what permissions I need to change.
>
> I've changed the owner on the data directories, ERLang directories any many more besides, but I'm still getting the error.
>
> If anyone could point me in the right direction, I'd be hugely grateful.
>
> Thanks,
>
> Martin
Most of the info is in here:
https://github.com/apache/couchdb/blob/master/INSTALL.Unix#L175-187
and some on the wiki
http://wiki.apache.org/couchdb/Installing_on_Ubuntu or similar ones
for other linux variants may help.
su to the couch user, and then run couchdb interactively `couchdb -i`
which sometimes spits out a moment of clarity.
IIRC there was a recent-ish example on the user@ mailing list but I
unfortunately don't recall the solution!
A+
Dave