I'm moving mailbox data from the old zimbra server (oldmail.domain.com) to the new zimbra server (newmail.domain.com) using Imapsync.
I use the following command:
imapsync --nofoldersizes --skipsize --fast --nosyncacls --syncinternaldates --host1 oldmail.domain.com --user1 user#domain.com --authuser1 admin --password1 password1 --ssl1 --host2 newmail.domain.com --user2 user#domain.com --authuser2 admin --password2 password2 --ssl2
Why when on a new server, mailbox data requires a lot of hard disk space? The old server is under 100GB, the new server is more than 200GB. Meanwhile, the mailbox data that is moved is the same. I check the inbox of each account, there are no duplicate emails.
Why did it happen? If there is a duplication, how do I check the duplication? And how do I delete the duplicated mailbox data?
maybe there are some saved searches or shared folders in your account?
Setelah bertanya ke beberap forum, akhirnya saya mendapatkan jawaban.
Imapsync reads email as a separate object. So if an email is sent and cc to 10 accounts, imapsync will store as many as 10 emails, thus causing greater storage consumption.
In zimbra, there is a zmdedupe command which is executed in zimbra user mode. This command will make these emails saved only once, and will create a link to access them. So that it can save more storage
#su - zimbra
$zmdedupe start
We can see the status of this process by typing the command
zmdedupe status
From this command too, we can know how much storage can be saved.
Related
I'm using mutt with PGP and would like to store sent messsages in a corresponding IMAP folder. Problem: When I send an encrypted email to somebody, mutt copies the encrypted email to the sent folder, instead of an unencrypted version. Of course, since it is encrypted with the receivers key, I cannot unencrypt this message anymore, making it useless.
Is there a setting in mutt to fix this?
The setting you ask for is set fcc_clear=yes. This will save it in cleartext to the Sent folder (see man muttrc for full explanation).
There was a question to the opposite on superuser, so I'll extend my answer to that for a full picture.
Depending on where your "Sent" folder resides, either in a remote IMAP server you want to have fulltext search for (probably need cleartext), or on a local disk that is encrypted at-rest, you have further options:
mutt currently has pgp_self_encrypt in tandem with pgp_default_key (added to mutt on 2017-05-29 / 2018-01-15, so check your version) to control the behaviour
gnupg has an encrypt-to <keyid> option, but I would prefer the mutt config
neomutt has encrypt-to-self
So i got thunderbird handeling all my mail. Have some extensions for copying and removing duplicated mails.
But now i got even larger mailboxes. They have about 100K mails with or without attachments.
And if i copy the mail from the server to the local folders i get the following message:
"The xxxx folder is full and can not contain any incoming messages. Delete old or unwanted mail and compact the folder to make room for more messages."
I thought this was an old message or a message if youre disk is formatted in fat32 (Can't handle files over 4gb)
So what limit am i reaching or what can i do to copy these folders from the server to the local folders.
I googled it but i keep going in cycles and can't find a solution.
What considerations should be addressed when moving a Microsoft 2003 server with MSMQ installed to another domain?
We assigned permissions to the queue for a user account from the new domain. We still get an error "Access to Message Queing system is denied."
There are no other servers in the new domain running MSMQ. Is there an Active Directory object that gets created when MSMQ is installed on a member server?
The issue has been resolved. The problem was with a private queue created by our application. That explains why we didn't see any AD object created. Using domain admin credentials, we tried to modify the permissions on the private queue using the MMC snapin. We got an error and could not make the required change because the domain admin was not on the ACL for that queue. We found a work around to the problem. We created a new temporary private queue and assigned full control to the user account we need for our application. We located the newly created private queue configuration file in the \system32\msmq\storage\lqs directory. We opened the file and copied the entire security line. Then we opened the configuration file for the private queue we are trying to fix and pasted the security line over the same line in that file. We saved the configuration file. We restarted the MSMQ service and then started our application service. The application could now function properly. Thanks everyone for your help!
If the domain is in a different forest and you wish to make use of public queues, etc. then you will need to reinstall MSMQ so that all the MSMQ objects are created in AD.
What exact operation generates "Access to Message Queing system is denied."?
If you don't have any messages or journal history that you need to save, I would suggest uninstalling and reinstalling MSMQ.
Have you logged into the server directly using the account that is trying to send the message? This is necessary to establish a message queuing certificate, a little understood and poorly documented area of MSMQ.
We also found another easier way to reset the permissions on the private queue. Instead of modifying the private queue configuration file, we took ownership of the private queue as the domain admin, assigned full permissions to the user account needed for our application, and transfered ownership to the same user. This restored the private queue to the state it was in prior to the server move.
I am having issues sending mail from my server, when I type in phpinfo() I get this:
Mail: /var/spool/mail/frank
...
_ENV["MAIL"]: /var/spool/mail/frank
The php file that I'm using the Mail function in is owned by Apache and has 775 permissions. Inside /var/spool/mail there is no apache file, but there is a file for every other user on the box. Do I need to change the php.ini? Or can I add in an apache file (perhaps duplicate the frank file and then rename)?
Any advice can help!
/var/spool/mail is where incoming unread mail is stored. There would be an 'apache' file only if the apache account has received mail.
Have you looked inside the MTA's log (usually /var/log/maillog) to see what's going on? If PHP's properly sending mail, there should be an entry in there for the mail being en-queued and what happened when the MTA (postfix/sendmail/exim/etc...) tries to forward it onwards.
I want to get the alert message for every execution of the scp command. If some one access my file using the scp command, that time I have to receive one mail. How can I achieve this. Please help me.
Unless you have administrative control of the server, what you ask is not possible -- Unix filesystems simply don't work that way. You can set your file permissions so that other users can't find or read your files, but there is no way for you, as an unprivileged user, to request email notification upon attempted or successful access attempts, via scp or any other method.
Are you on the server, or client side?
server side: Key-based ssh authentication has the ability to start custom commands when the user log into the system (can by bypassed if the user can change the authorized_keys file). There you can add a hook to send a mail on the login, or start a own sftp server with the desired logging enabled.
The right way to do that is to replace the scp program on the server by your own version (or by a wrapper).