Page 1 of 2 12 LastLast
Results 1 to 10 of 19

Thread: Zimbra 8 Migration from Scalix to Zimbra - Split Domain

  1. #1
    hrobinson is offline Intermediate Member
    Join Date
    Mar 2013
    Posts
    21
    Rep Power
    2

    Default Zimbra 8 Migration from Scalix to Zimbra - Split Domain

    Hello all,

    I am planning a migration from Scalix to Zimbra 8.0.2. For a time, not more than a month or two, I want to have both Scalix, and Zimbra online as I migrate Email accounts from Scalix to Zimbra. This is a small domain of 15 users. Some accounts are huge (36+ gigs) and could take a weekend to migrate.

    An outline to my plan might look like this:
    1. select a Victim -- I mean user.
    2. Notify victim (um ... user) of the impending migration to the new server.
    3. Set up the new account on the Zimbra server.
    4. After the victim (yea ... user) leaves the office for the evening, Migrate the Email from the Scalix server to the Zimbra server.
    5. Disable the Victims (I mean Users )account on the Scalix server.
    6. Test the sending and receiving of Email on the new Zimbra server.
    7. Remove old Scalix connector from laptop.
    8. Install new Zimbra connector on laptop
    9. Configure software as necessary.
    10. If the Victim (I really mean user) has a smart phone, adjust the settings on that device also. Most have Android, some have iPhone or iPad.
    11. Wash Rinse and Repeat until all users are converted.

    My questions:

    How does the Scalix server forward Email to the Zimbra server? I have found "Split Domain" in the Zimbra Wiki that I will read, but I was hoping someone would have some additional ideas for me.

    Once all Email has been migrated, how do I make sure after I shut down Scalix that the Zimbra server will not attempt to forward email back to the decommissioned Scalix server?

    I am putting together all the things I need to do when I deliver the server to the office so I may have more questions relating to this later.

    Thanks for your help.

    -- Harold

  2. #2
    LMStone's Avatar
    LMStone is offline Moderator
    Join Date
    Sep 2006
    Location
    477 Congress Street | Portland, ME 04101
    Posts
    1,373
    Rep Power
    10

    Default

    Quote Originally Posted by hrobinson View Post
    Hello all,

    I am planning a migration from Scalix to Zimbra 8.0.2. For a time, not more than a month or two, I want to have both Scalix, and Zimbra online as I migrate Email accounts from Scalix to Zimbra. This is a small domain of 15 users. Some accounts are huge (36+ gigs) and could take a weekend to migrate.

    An outline to my plan might look like this:
    1. select a Victim -- I mean user.
    2. Notify victim (um ... user) of the impending migration to the new server.
    3. Set up the new account on the Zimbra server.
    4. After the victim (yea ... user) leaves the office for the evening, Migrate the Email from the Scalix server to the Zimbra server.
    5. Disable the Victims (I mean Users )account on the Scalix server.
    6. Test the sending and receiving of Email on the new Zimbra server.
    7. Remove old Scalix connector from laptop.
    8. Install new Zimbra connector on laptop
    9. Configure software as necessary.
    10. If the Victim (I really mean user) has a smart phone, adjust the settings on that device also. Most have Android, some have iPhone or iPad.
    11. Wash Rinse and Repeat until all users are converted.

    My questions:

    How does the Scalix server forward Email to the Zimbra server? I have found "Split Domain" in the Zimbra Wiki that I will read, but I was hoping someone would have some additional ideas for me.

    Once all Email has been migrated, how do I make sure after I shut down Scalix that the Zimbra server will not attempt to forward email back to the decommissioned Scalix server?

    I am putting together all the things I need to do when I deliver the server to the office so I may have more questions relating to this later.

    Thanks for your help.

    -- Harold
    Hi Harold,

    The much easier way to do this is to follow the Split Domain Wiki article but make Zimbra the Primary. Essentially you will:
    1. Provision all the mailboxes on the Zimbra server and deploy Split DNS on the Zimbra server to get all the private DNS records correct.
    2. Change the zimbraMailTransport attribute for each mailbox to point to your Scalix server; test sending email from User A to User B and make sure it ends up in User B's mailbox in Scalix. If not, you've done something wrong and need to fix it before proceeding further.
    3. Change public DNS so all inbound email flows through Zimbra.
    4. Use imapsync or a tool of your choice to migrate emails from Scalix to Zimbra.
    5. Set a cutover date, and on that date change the zimbraMailTransport attribute back to the default. You can do this on a per-mailbox basis to migrate users one or a few at a time.
    6. Have the users stick to the Zimbra web interface if at all possible. The Zimbra Connector for Outlook installs a new Outlook profile; users who don't know how to switch between Outlook profiles may get annoyed that their Scalix email has gone away so abruptly.


    At that point you should be able to retire the Scalix server.

    The tricky bit is private DNS; Zimbra is very sensitive to this being set up correctly.

    Hope that helps,
    Mark

    WebRep


    currentVote


    noRating
    noWeight

  3. #3
    hrobinson is offline Intermediate Member
    Join Date
    Mar 2013
    Posts
    21
    Rep Power
    2

    Default

    LMStone,

    Thank you for your response. This all does make sense.

    A few of thoughts:

    1. Provisioning all mail boxes at once makes absolute sense.
    Q: I believe I already employ a split DNS at the office. When outside the office mail.sbgk.com points to the WAN internet address. when in the office mail.sbgk.com points to the internal server on the 10.x.x.x network. If that is the case. Should I turn off BIND on the Zimbra server?

    2. Thanks for the ZimbraMailTransport attribute. I will look at this shortly to understand how it works.

    3. Our public DNS points to our WAN IP address. I use NAT and forward all port 25 traffic to the correct server on our LAN. So when I need to switch over, I will just adjust Port Forwarding on the firewall to point to the new Zimbra server when I am ready.
    Q: Also all our Email is pre-filtered for viruses and spam through a service called Mailroute before it arrives onsite. Thoughts on that?

    Note: I have port 25 restricted to our Scalix server. No other machines with the exception of the copiers are allowed to use port 25. I will need to adjust the firewall so that it will allow the new Zimbra server to communicate on port 25 also.

    4. ImapSync tool will be used to move mail. Unfortunately it does not move calendar items also (from what I have seen so far) so my understanding is that I will have to export all the calendar contacts to PST then import them into Zimbra.
    Q: Is this correct? Do you have other suggestions that might work better?

    5. I have decided that people will not be switching back and forth from Zimbra to Scalix. When I tell them that I am moving their mail, they will not have access to the old mail server. I may possibly create a virtual machine to allow them access to the old scailx server for a small time. I have not thought about that aspect of the project yet.

    6. The office manager has yet to see the web interface. They are so use to Outlook, that I doubt they will move to the web interface. I really wish they would. Never the less I think they will be really impressed with the new interface vs Scalix web interface which was problematic to access if it worked at all.

    All of these suggestions have been really helpful. <sigh> more reading to do. Such is the life of a sysadmin ;-)

    Thanks again for your help

    Sincerely,

    -- Harold

  4. #4
    LMStone's Avatar
    LMStone is offline Moderator
    Join Date
    Sep 2006
    Location
    477 Congress Street | Portland, ME 04101
    Posts
    1,373
    Rep Power
    10

    Default

    Hi Harold,

    Sounds like you are making progress on your plan!

    To answer your questions:

    DNS: If you already have a internal-only DNS server, sure, you can use that instead of BIND on the Zimbra server. Zimbra is very sensitive to DNS being set up correctly, so please be sure to follow the installation guide and run all the troubleshooting DNS checks.

    Port Forwarding: Sounds like you have a mail gateway appliance for both inbound and outbound email? If so, and assuming the gateway appliance is the public MX for your domain, then yes as part of the cutover just do port 25 TCP forwarding to Zimbra instead of to Scalix. You may also want to forward TCP 587 so Outlook users on the road can authenticate and send email through the Zimbra server encrypted.

    ImapSync: Yes, this tool only does mail, but can be run multiple times without creating duplicate emails. It also recreates the email folder tree and we have found it to be excellent. For calendars and contacts, yes, just export them from Scalix into new PST files and then use the Zimbra PST Import Wizard tool to do the imports.

    Please keep us posted as to your progress.

    All the best,
    Mark

  5. #5
    hrobinson is offline Intermediate Member
    Join Date
    Mar 2013
    Posts
    21
    Rep Power
    2

    Default

    Mr. Stone,

    Thanks a gain for replying to my questions. Today was the day I delivered the Zimbra server for testing to my client. This may consist of several posts as I progress through installation at the clients site.

    Prior to transporting the server from my lab to the clients site, I stopped the named service and prevented it from starting automatically as I will not need it at the client's site. As this is installed on CentOS 6.4 64 bit with 8GB of RAM, I was lazy and used the setup command. Went into services and press the space bar next to named. Next I issued the following commands:

    #service named stop

    Next I shut down zimbra by issuing the following commands:

    #su zimbra
    $zmcontrol stop
    $exit
    #halt


    This stopped the Zimbra services and also prevents them from starting up when the server restarts. Even though the Zimbra services are shut down properly when issuing the halt command, I am going to continue to shut them down first before restarting the server. For me this feels safer.

    Then I packed up the server and traveled to the clients office. Once there I plugged the server into the UPS, connected to the network. At this point the switch at the clients office is completely full. I have also installed a new wireless router and had 4 free gigabit ports available so I plugged the Zimbra server into one of the free ports. This was a mistake that I dicovered early on. After assigning a free IP address to the NIC card, I was puzzled as to why I could not ping it. Everything looked ok. I decided to change the NIC on the Zimbra server so that it was DHCP. This revealed that I was on the wrong network (10.1.2.x not 10.1.3.x). Darn, so much for the easy way out. I ended up locating a free port on the switch. I changed the ip address back to the .3.x network and the server started answering pings.

    Next I updated our DNS server so that it knows about our new Zimbra server. I know this is old school, but I built this DNS server by hand and have not had any problems with it since its implementation back in 2000. I also have a secondary backup DNS server on the firewall. (Its part of the Smoothwall services) so I also added zimbra to this server through the web interface. I stopped and started the named services and tested with the following command:

    #dig sbgk.com mx

    but dig did not find my mx record, nor did it find zimbra, nor zimbra.sbgk.com. Puzzled I examined my system log to see what was wrong. This is what I found:

    Mar 12 12:33:20 sbgk-2 named[746]: starting BIND 9.2.4 -u named -t /var/named/chroot
    Mar 12 12:33:20 sbgk-2 named[746]: using 1 CPU
    Mar 12 12:33:20 sbgk-2 named[746]: loading configuration from '/etc/named.conf'
    Mar 12 12:33:20 sbgk-2 named[746]: listening on IPv4 interface lo, 127.0.0.1#53
    Mar 12 12:33:20 sbgk-2 named[746]: listening on IPv4 interface eth0, 10.1.3.2#53
    Mar 12 12:33:20 sbgk-2 named[746]: command channel listening on 127.0.0.1#953
    Mar 12 12:33:20 sbgk-2 named[746]: zone 0.in-addr.arpa/IN: loaded serial 42
    Mar 12 12:33:20 sbgk-2 named[746]: zone 3.1.10.in-addr.arpa/IN: journal rollforward failed: journal out of sync with zone
    Mar 12 12:33:20 sbgk-2 named[746]: zone 0.0.127.in-addr.arpa/IN: loaded serial 1997022700
    Mar 12 12:33:20 sbgk-2 named[746]: zone 255.in-addr.arpa/IN: loaded serial 42
    Mar 12 12:33:20 sbgk-2 named[746]: zone 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0. 0.0.0.0.0.0.ip6.arpa/IN: loaded serial 1997022700
    Mar 12 12:33:20 sbgk-2 named[746]: zone sbgk.com/IN: journal rollforward failed: journal out of sync with zone
    Mar 12 12:33:20 sbgk-2 named[746]: zone localdomain/IN: loaded serial 42
    Mar 12 12:33:20 sbgk-2 named[746]: zone localhost/IN: loaded serial 42
    Mar 12 12:33:20 sbgk-2 named: named startup succeeded
    Mar 12 12:33:20 sbgk-2 named[746]: running


    Googling this error brought me to this answer:[BIND] "journal rollforward failed: journal out of sync with zone" ERROR

    I removed the .jnl files for both zones (sbgk.com and 3.1.10.in-addr.arpa) and restarted bind. Bind then loaded correctly and allowed me to start resolving the zimbra name, as seen in this dig example:

    Mar 12 14:29:59 sbgk-2 named[1671]: starting BIND 9.2.4 -u named -t /var/named/chroot
    Mar 12 14:29:59 sbgk-2 named[1671]: using 1 CPU
    Mar 12 14:29:59 sbgk-2 named[1671]: loading configuration from '/etc/named.conf'
    Mar 12 14:29:59 sbgk-2 named[1671]: listening on IPv4 interface lo, 127.0.0.1#53
    Mar 12 14:29:59 sbgk-2 named[1671]: listening on IPv4 interface eth0, 10.1.3.2#53
    Mar 12 14:29:59 sbgk-2 named[1671]: command channel listening on 127.0.0.1#953
    Mar 12 14:29:59 sbgk-2 named[1671]: zone 0.in-addr.arpa/IN: loaded serial 42
    Mar 12 14:29:59 sbgk-2 named[1671]: zone 3.1.10.in-addr.arpa/IN: loaded serial 2013031201
    Mar 12 14:29:59 sbgk-2 named[1671]: zone 0.0.127.in-addr.arpa/IN: loaded serial 1997022700
    Mar 12 14:29:59 sbgk-2 named[1671]: zone 255.in-addr.arpa/IN: loaded serial 42
    Mar 12 14:29:59 sbgk-2 named[1671]: zone 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0. 0.0.0.0.0.0.ip6.arpa/IN: loaded serial 1997022700
    Mar 12 14:29:59 sbgk-2 named[1671]: zone sbgk.com/IN: loaded serial 2013031201
    Mar 12 14:29:59 sbgk-2 named[1671]: zone localdomain/IN: loaded serial 42
    Mar 12 14:29:59 sbgk-2 named[1671]: zone localhost/IN: loaded serial 42
    Mar 12 14:29:59 sbgk-2 named: named startup succeeded
    Mar 12 14:29:59 sbgk-2 named[1671]: running
    Mar 12 14:29:59 sbgk-2 named[1671]: zone 3.1.10.in-addr.arpa/IN: sending notifies (serial 2013031201)
    Mar 12 14:29:59 sbgk-2 named[1671]: zone sbgk.com/IN: sending notifies (serial 2013031201)
    Mar 12 14:29:59 sbgk-2 named[1671]: received notify for zone '3.1.10.in-addr.arpa'


    As you can see above the named service is now running properly with no errors. and when I issue the dig command now I have the following result:

    #dig sbgk.com mx
    ; <<>> DiG 9.8.2rc1-RedHat-9.8.2-0.10.rc1.el6_3.6 <<>> sbgk.com mx
    ;; global options: +cmd
    ;; Got answer:
    ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 25964
    ;; flags: qr aa rd ra; QUERY: 1, ANSWER: 2, AUTHORITY: 1, ADDITIONAL: 3

    ;; QUESTION SECTION:
    ;sbgk.com. IN MX

    ;; ANSWER SECTION:
    sbgk.com. 86400 IN MX 10 grumpy.sbgk.com.
    sbgk.com. 86400 IN MX 10 zimbra.sbgk.com.

    ;; AUTHORITY SECTION:
    sbgk.com. 86400 IN NS www.sbgk.com.

    ;; ADDITIONAL SECTION:
    grumpy.sbgk.com. 86400 IN A 10.1.3.6
    zimbra.sbgk.com. 86400 IN A 10.1.3.5
    www.sbgk.com. 86400 IN A 216.230.216.126

    ;; Query time: 2 msec
    ;; SERVER: 10.1.3.2#53(10.1.3.2)
    ;; WHEN: Tue Mar 12 14:03:47 2013
    ;; MSG SIZE rcvd: 138

    [root@zimbra ~]#


    I know this is successful by looking at the line "flags" It shows me there are two answers. You will also notice the old server is named "grumpy.sbgk.com" The server is named this not because I am a Disney fan, but because when this server goes down the whole office is "grumpy"

    So it looks like everything is just fantastic with the network and the new Zimbra server. so I issue the following command:

    su zimbra
    zmcontrol start
    [zimbra@zimbra root]$ zmcontrol start
    Host zimbra.sbgk.com
    Starting ldap...Done.
    Failed.
    Failed to start slapd. Attempting debug start to determine error.
    513f8b7e daemon: bind(7) failed errno=99 (Cannot assign requested address)
    513f8b7e slap_open_listener: failed on ldap://zimbra.sbgk.com:389


    What the heck? everything should be ok. so back to google to see what it has to say about this problem: I come up with a Zimbra forum post which I will point to here: http://www.zimbra.com/forums/install...art-slapd.html.

    I thought about this for a second. then realized that the hosts file had not been updated with the new IP address. It still had the old IP address from the lab. I went into the hosts file and updated it with the correct IP address then restarted Zimbra again.

    [zimbra@zimbra etc]$ zmcontrol start
    Host zimbra.sbgk.com
    Starting ldap...Done.
    Starting zmconfigd...Done.
    Starting logger...Done.
    Starting convertd...Done.
    Starting mailbox...Done.
    Starting antispam...Done.
    Starting antivirus...Done.
    Starting snmp...Done.
    Starting spell...Done.
    Starting mta...Done.
    Starting stats...Done.
    [zimbra@zimbra etc]$


    Wahoo, looks like we are off to the races. Now I need to add all the users from the old server. I will do this by hand as there is only 15 users.

    Keep a watch on this thread as I post my experiences with installing Zimbra with an already running Email server (sacalix).

    --Harold Robinson

  6. #6
    hrobinson is offline Intermediate Member
    Join Date
    Mar 2013
    Posts
    21
    Rep Power
    2

    Default

    All,

    Today is the day that I redirect the flow of Email from the Scalix server to the Zimbra server.

    What I did to accomplish this was to manually enter all user information. Since there is only 16 accounts, I did this manually. Any larger than 25, and I would have extracted all the names from Scalix and did a import from a text file of some kind.

    With all the Email accounts setup and passwords properly set I then used this command to route all mail from zimbra to Sclix:

    zmprov ma firstname.lastname@sbgk.com zimbraMailTransport smtp:scalixmailserver.sbgk.com:25

    Next I sent a test Email from the Zimbra web interface to all the email accounts setup on zimbra. This email was then automatically routed to the old Scalix server. Every user replied that they received the Email. This means that I can tell the firewall to route all mail to the new zimbra mail server.

    -------

    Before I do that, the new Zimbra server update was released. Since the server is not yet production, I felt it necessary to update the server before directing mail flow to the new zimbra server. I started going through the zimbra update process and it told me a full backup had not been performed in the last 24 hours. I have no data on this server yet, but while I was testing the server I wanted to make sure everything was working correctly.

    First I checked the home screen, The web interface is telling me I have 53 days left on my license. I also see under maintenace the last backup was performed March 13, 2013. Hmmm checking my watch that was yesterday. I was thinking that maybe the backup has to be very current before doing an upgrade. This makes sense to me.

    So I went to the command prompt and executed the following commands:

    su zimbra
    zmbackup -f -a all zimbra.sbgk.com (This is the command to issue a full backup to the zimbra server. I found this in the Wiki.)


    After a pause I received this really odd error:

    Error occured: License is not valid or feature AccountsLimit is disabled

    Uh? What is this? I started googleing around, did not find anything relating to this. Attempted to call zimbra support, and encountered the chicken before the egg scenario where I need to have access to the support portal before calling them. Since I am trying the software out, I have not purchased support for the product. Therefor I don't have access to support portal.

    Some nice people at Vmware supplied me with several numbers. One of them ended me up back at the same place (the 888 number for zimbra support) and no access to a warm body to talk to. Pressing 0 simply dumped me into a fast busy signal. The other was the number to Thomas, the support escalation (manager?) I don't know if Vmware was to give out this poor persons phone number but I have it never the less. She was so nice she gave me two number. The first went to the avaya email system. The other was a direct line to his voice mail. I left a really nice message explaining my situation and hung up (leaving my phone number of course).

    I was a little frustrated that I was having this problem and no way to resolve it quickly. My instincts told me this was really simple. I tried google again using different search terms and found this article from the Zimbra forums that was similar to the problem I was experiencing but not exactly the same. Here is the article that ultimately solved my problem: License Error: Mail: AccountsLimit exceed: 2/1

    You see when an administrator installs the zimbra server software for the first time, after the installation has been completed, you must ACTIVATE the license. There was no warning on the administration console home page to tell you it requires activation. It just simply tells you how many days you have left. on a hunch on the right sized of the admin console there is a Server Status box. In that box it tells me:

    Your Trial license expiration date is Monday, May 6, 2013 3:24:05pm

    Below that is a link to manage licenses. I click on that and I am transferred to the License information page. It tells me the number of licenses, activation ID, and product version. Right below that in RED letters it tells me "Your License has not been activated. Activate the license from the tool bar!"

    OK ... hmmm what tool bar is the server talking about? I look all over the place and can't find an activate button. I went back to the article mentioned above which did not help me much. I goggled around and found that if you click on the gear at the top where it says Help, Save, Close. The gear is to the right of Close, then click "Activate License" the trial license will activate and the red letters will go away.

    Now I try this command again: zmbackup -f -a all -s myzimbraserver.sbgk.com I can now see that I have a full backup.

    As a side note. This server was going to be used as a vmware sever with Zimbra running on top of it. Unfortunately the hardware is too old to handle vmware 5.1. That was the bad news. The good news is that this server actually has two raid 5 arrays setup. I selected the largest array to be the zimbra datastore. This array is mounted on /opt.

    The other array is mounted at /Data/DS-1. It is here that I store the zimbra backups. I already reconfigured zimbra to store all backups here. If it gets full (which it should not because I configured backups to not run if 10% or less space is available) it will not affect the Zimbra datastore. Then I can clean out the backups if necessary and have backups run again. All with out affecting the space of the Zimbra Datastore.

    As I can see now, I have a "full-20130314.230015.836" file that appeared after the backup completed. Now I will try to run the upgrade.

    ---------

    1. After unpacking 8.0.3 and going into the folder and running ./install.sh --platform-override the script starts its pre-flight check.

    2. The install script then asks me to accept the license agreement. I hit Y three times. Prior to this it would tell me that it could not find a backup and if I wanted to proceed with out one. Now it is asking me if i want to verify message store database integrity. I hit Y and hit enter

    3. after no errors were found, the installation script asks me if I want to upgrade. I hit Y to that question and it then asks me do I want to install zimbra-memcached or zimbra-proxy. I answer N to both. Then I answer Y to install anyway? Since this server is Cent-OS 6.3 which is exactly the same as RedHat REL 6.3 I can use the platform override option safely. next I answer Y to the question "The system will be modified. Continue?" After this it starts shutting down the server and updating the Zimbra software

    4. The install script now tells me "Configuration complete - press return to exit". All looks good so far. lets log in an see what happens! Well no smoke from the server so everything looks good here. If I go to "Manage Licenses" and look at the "Product Version" I see that the version is now 8.0.3_GA_5664. So from a zimbra stand point I am ready to go.

    ----------

    After the zimbra upgrade has completed I also see that I need to update CentOS from 6.3 to 6.4. I will do this now before I convert any user accounts to Zimbra. That way the server is absolutely up-to-date before we really start testing. To update from CentOS 6.3 to 6.4 just use "yum update". If you are moving from 32 bit to 64 bit, you will need to re-install the operating system. It really is not necessary to do this OS update right now. I am sure 6.3 is just fine. I am doing this for the experience. Also At this point in the project if I break anything I loose no data, but I gain valuable experience. So I am going to take advantage of the opportunity.

    1. I will issue the following command which will stop zimbra mail services prior to the CentOS OS upgrade.
    su zimbra
    zmcontrol stop

    Zimbra responds with:
    Stopping vmware-ha...Done.
    Stopping zmconfigd...Done.
    Stopping stats...Done.
    Stopping mta...Done.
    Stopping spell...Done.
    Stopping snmp...Done.
    Stopping cbpolicyd...Done.
    Stopping archiving...Done.
    Stopping antivirus...Done.
    Stopping antispam...Done.
    Stopping proxy...Done.
    Stopping memcached...Done.
    Stopping mailbox...Done.
    Stopping convertd...Done.
    Stopping logger...Done.
    Stopping ldap...Done.

    2. good next I exit out of zimbra
    $exit
    exit
    [root@zimbra~]#


    3.next I issue the update command for cent-os
    [root@zimbra~]#yum update

    4. yum running under CentOS 6.3 reports that it will install 20 new packages and upgrade 326 packages. The total download will be 288Megs. I answer Y to the question and yum starts downloading the necessary packages.

    5. After the upgrade has completed (takes about 45 minutes or so) I reboot the CentOS server.

    6. After a successful reboot, I start the Zimbra Email Services by issuing the following command
    su zimbra
    zmcontrol start


    7. Zimbra responds with:
    Host zimbra.sbgk.com
    Starting zmconfigd...Done.
    Starting logger...Done.
    Starting convertd...Done.
    Starting mailbox...Done.
    Starting antispam...Done.
    Starting antivirus...Done.
    Starting snmp...Done.
    Starting spell...Done.
    Starting mta...Done.
    Starting stats...Done.
    [zimbra@zimbra root]$


    8. Well it looks like all upgrades were performed successfully. Now on to selecting a victim ... I mean user (mainly myself) to convert Email over to the Zimbra Server.

    Sincerely,

    Harold Robinson

  7. #7
    hrobinson is offline Intermediate Member
    Join Date
    Mar 2013
    Posts
    21
    Rep Power
    2

    Default zimbraBatchedIndexingSize for Zimbra 8

    All,

    I am in the process of writing a script that will convert my users mail over to Zimbra.

    A lot of articles state that we should be setting zimbraBatchedIndexingSize 20 to use the batched index feature of Zimbra.

    1. Several articles refer to making this a default option in future versions of Zimbra. As per this Zimbra wiki article: Performance Tuning Guidelines for Large Deployments - Zimbra :: Wiki as of ZCS 5.0.3 this was a bug and you needed to run a script to turn this option on, to ensure proper performance. Do I still need to turn this option on for Zimbra 8? Is this option per Email account?

    Thanks for all your help.

    Sincerely,

    Harold Robinson

  8. #8
    phoenix is online now Zimbra Consultant & Moderator
    Join Date
    Sep 2005
    Location
    Vannes, France
    Posts
    23,566
    Rep Power
    57

    Default

    Quote Originally Posted by hrobinson View Post
    Do I still need to turn this option on for Zimbra 8?
    The easiest way to find that out would be to check the bug report for a date that it was fixed and whether it will be set as the default in a future release - it is for new installs. You can also check what the setting is by using zmprov to query your server.

    Quote Originally Posted by hrobinson View Post
    Is this option per Email account?
    As per the article to which you've linked, the details in the little script show it as being a COS attribute and as I mentioned above you can quest the COS with zmprov to determine the setting.
    Regards


    Bill


    Acompli: A new adventure for Co-Founder KevinH.

  9. #9
    hrobinson is offline Intermediate Member
    Join Date
    Mar 2013
    Posts
    21
    Rep Power
    2

    Default

    Quote Originally Posted by phoenix View Post
    The easiest way to find that out would be to check the bug report for a date that it was fixed and whether it will be set as the default in a future release - it is for new installs. You can also check what the setting is by using zmprov to query your server.

    As per the article to which you've linked, the details in the little script show it as being a COS attribute and as I mentioned above you can quest the COS with zmprov to determine the setting.
    Thanks for your reply,

    Unfortunately this does not answer my question. COS is what? Class Of Service?, Cost of Service? I am a little confused. What COS do I put in the parameter? or maybe I am asking the wrong questions

    ... but I will really look at that script to see if I can figure this out. Being a new administrator to Zimbra it does take a LOT of reading to catch up.

    Sincerely,

    Harold Robinson

  10. #10
    phoenix is online now Zimbra Consultant & Moderator
    Join Date
    Sep 2005
    Location
    Vannes, France
    Posts
    23,566
    Rep Power
    57

    Default

    Quote Originally Posted by hrobinson View Post
    Unfortunately this does not answer my question. COS is what? Class Of Service?, Cost of Service? I am a little confused. What COS do I put in the parameter? or maybe I am asking the wrong questions
    It's Closs of Service and you (might) need to apply to whichever COS (or all of them) you have in use, as I mentioned you should verify if the attribute is set.

    Quote Originally Posted by hrobinson View Post
    ... but I will really look at that script to see if I can figure this out.
    It's a fairly simple command, the one in the article modifies the COS and you want to get the attribute first for verification, post again if you still have a problem.
    Regards


    Bill


    Acompli: A new adventure for Co-Founder KevinH.

Page 1 of 2 12 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. Replies: 2
    Last Post: 07-26-2012, 09:36 AM
  2. Scalix to Zimbra Migration
    By chasecrum in forum Migration
    Replies: 0
    Last Post: 06-09-2011, 01:50 PM
  3. Split Domain Migration
    By cybil in forum Migration
    Replies: 8
    Last Post: 04-02-2010, 12:32 PM
  4. Split domain migration
    By zimkeeps in forum Migration
    Replies: 0
    Last Post: 05-06-2008, 09:55 PM
  5. Replies: 1
    Last Post: 10-26-2007, 08:21 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •