I can not run mu index when mu4e/mu-server about mu HOT 11 CLOSED

I assume that if mu server is running then Emacs is also running (mu server is started by mu4e) and Emacs server is enabled. And my way to solve the problem is the following.

#!/usr/bin/env bash
if pgrep -f 'mu server'; then
    emacsclient -e '(mu4e-update-index)'
else
    mu index
fi

from mu.

Just a tip: if you have an Emacs daemon running, you can index with mu4e using the following shell command:

emacsclient -e '(mu4e-update-index)'

from mu.

Just a important note for the readers:

The full command set to it work fine is:

pkill -2 -u $UID mu
sleep 1
mu index

The 1 second delay was needed for me. ;-)

from mu.

Remove the ~/.mu and re-run mu index.

from mu.

The underlying database (Xapian) can be opened in read/write-mode only once... To prevent this error, you can either retrieve your mail from within mu4e (ie., 'U' in the main view), or in your cron script put something like:

pkill -2 -u $UID mu

This will sent a SIGINT to the mu process, which will then terminate gracefully. Mu4e will automatically restart mu when it is needed again (of course, you can still only have one mu at the same time, but in practice it works quite well).

from mu.

Hi Dirk-Jan,

I have a similar problem, and of course your answer clarify everything. But my setup is a little bit different:
I make offlineimap+mu run in background. But when I launch mu4e the lock is taken and mu update does not work anymore. I have to update the db manually, which is not nice (make the email reading not interactive). Is it possible to force mu4e to open the xapian database read only, and let the offlineimap daemon to run mu index?

Thank you.

from mu.

That's not possible...mu4e cannot do its job when it cannot write the database.

However, what you can do is either the pkill trick mentioned above, or in your background script check for a running mu and if there is already, let mu4e handle indexing.

from mu.

See also the related discussion in the list: https://groups.google.com/forum/#!searchin/mu-discuss/Keybinding$20for$20mu4e-update-index|sort:relevance/mu-discuss/MCbP8Gwiujc/kg1QA2rPAgAJ

from mu.

I was doing what @tituomin writes and from version 1.x of mu I cannot even run mu4e on Emacs.

error in process sentinel: mu4e~proc-sentinel: Database is locked by another process
error in process sentinel: Database is locked by another process

Running pkill -2 -u $UID mu on a console and then trying again yields the same result.

I swear I haven't touch my configuration of mu for Emacs or the configuration of my goimapnotify (nor touched the source code in months) and after the update to 1.x mu have been showing such behavior.

from mu.

Any idea how to suppress the message resulting from mu4e-update-index?
For some reason, (let ((inhibit-message t)) (mu4e-update-index)) still shows it.

I assume that if mu server is running then Emacs is also running (mu server is started by mu4e) and Emacs server is enabled. And my way to solve the problem is the following.

#!/usr/bin/env bash
if pgrep -f 'mu server'; then
    emacsclient -e '(mu4e-update-index)'
else
    mu index

from mu.

I assume that if mu server is running then Emacs is also running (mu server is started by mu4e) and Emacs server is enabled. And my way to solve the problem is the following.

#!/usr/bin/env bash
if pgrep -f 'mu server'; then
    emacsclient -e '(mu4e-update-index)'
else
    mu index
fi

Thank you,

but consider there is more than one person using mu4e simultaneously.
Then you have to check for "mu server" UID for this specific user.

Martin.

from mu.