This is what my understanding. May be I am also wrong here. 

Backup option is there in my new server, production2. 

When I use the command with -R option from new server, it is connecting to old server, Production1 and giving the output of the doveadm backup command in old server to get the data. If you see the command and error I posted, it shows the same. 

Sorry just incase if I am wrong.

Kishore

On Mon, 27 Jul, 2020, 11:32 pm Antonio Leding, <tech@leding.net> wrote:
Well that option set still doesn’t look correct…

Here is the doveadm option set from my 2.2.2 server which is several weeks\months before 2.2.36…



[Jul-27 @ 15:25:09 ] > dovecot --version
2.2.2

[Jul-27 @ 15:26:53 ] > ls -l /usr/local/bin/doveadm
-rwxr-xr-x 1 root root 1379203 May 22  2013 /usr/local/bin/doveadm

[Jul-27 @ 15:25:12 ] > doveadm
usage: doveadm [-Dv] [-f <formatter>] <command> [<args>]

  altmove      [-u <user>|-A] [-S <socket_path>] [-r] <search query>
  auth         cache|test
  backup       [-u <user>|-A] [-S <socket_path>] [-dfR] [-l <secs>] [-m <mailbox>] [-n <namespace>] [-s <state>] <dest>
  batch        [-u <user>|-A] [-S <socket_path>] <sep> <cmd1> [<sep> <cmd2> [..]]
  config       [doveconf parameters]
  copy         [-u <user>|-A] [-S <socket_path>] <destination> [user <source user>] <search query>
  director     add|dump|flush|map|move|remove|ring|ring|ring|status
  dump         [-t <type>] <path>
  exec         <binary> [binary parameters]
  expunge      [-u <user>|-A] [-S <socket_path>] [-d] <search query>
  fetch        [-u <user>|-A] [-S <socket_path>] <fields> <search query>
  force-resync [-u <user>|-A] [-S <socket_path>] <mailbox mask>
  fs           copy|delete|get|iter|iter-dirs|put|stat
  help         <cmd>
  import       [-u <user>|-A] [-S <socket_path>] [-s] <source mail location> <dest parent mailbox> <search query>
  index        [-u <user>|-A] [-S <socket_path>] [-q] [-n <max recent>] <mailbox mask>
  instance     list|remove
  kick         [-a <anvil socket path>] [-f] <user mask>[|]<ip/bits>
  log          errors|find|reopen|test
  mailbox      create|delete|list|mutf7|rename|status|subscribe|unsubscribe
  mount        add|list|remove
  move         [-u <user>|-A] [-S <socket_path>] <destination> [user <source user>] <search query>
  penalty      [-a <anvil socket path>] [<ip/bits>]
  proxy        kick|list
  purge        [-u <user>|-A] [-S <socket_path>]
  pw           [-l] [-p plaintext] [-r rounds] [-s scheme] [-t hash] [-u user] [-V]
  reload
  replicator   remove|replicate|status
  search       [-u <user>|-A] [-S <socket_path>] <search query>
  sis          deduplicate|find
  stats        dump|top
  stop
  sync         [-u <user>|-A] [-S <socket_path>] [-dfR] [-l <secs>] [-m <mailbox>] [-n <namespace>] [-s <state>] <dest>
  user         [-a <userdb socket path>] [-x <auth info>] [-f field] [-u] <user mask> [...]
  who          [-a <anvil socket path>] [-1] [<user mask>] [<ip/bits>]
  zlibconnect  <host> [<port>]


On Jul 27, 2020, at 3:28 PM, Kishore Potnuru <kishore.reachme@gmail.com> wrote:

Oh, I am extremely sorry. I was confused when typing the issue. Please help me in fixing the issue. Production1 is my old server. Production2 is my new server. Typing here correctly. Ignore all previous one's.

The new version is:
====================
[root@production2 ~]# dovecot --version
2.2.36 (1f10bfa63)


The old version is : 
======================
[root@production1 ~]# dovecot --version
2.0.9


I have executed the command from the new server:



=================================
[root@production2 test.org.com]# doveadm backup -R -u kishore@test.org.com remote:vmail@production1.baplc.com
vmail@production1.baplc.com's password:
usage: doveadm [-Dv] [-f <formatter>] <command> [<args>]

  altmove      [-u <user>|-A] [-S <socket_path>] <search query>
  auth         [-a <auth socket path>] [-x <auth info>] <user> [<password>]
  config       [doveconf parameters]
  director     add|flush|map|remove|status
  dump         [-t <type>] <path>
  expunge      [-u <user>|-A] [-S <socket_path>] <search query>
  fetch        [-u <user>|-A] [-S <socket_path>] <fields> <search query>
  force-resync [-u <user>|-A] [-S <socket_path>] <mailbox>
  help         <cmd>
  import       [-u <user>|-A] [-S <socket_path>] <source mail location> <dest parent mailbox> <search query>
  kick         [-a <anvil socket path>] [-f] <user mask>[|]<ip/bits>
  log          find|reopen|test
  mailbox      create|delete|list|mutf7|rename|status|subscribe|unsubscribe
  penalty      [-a <anvil socket path>] [<ip/bits>]
  purge        [-u <user>|-A] [-S <socket_path>]
  pw           [-l] [-p plaintext] [-r rounds] [-s scheme] [-u user] [-V]
  reload
  search       [-u <user>|-A] [-S <socket_path>] <search query>
  sis          deduplicate|find
  stop
  user         [-a <userdb socket path>] [-x <auth info>] <user mask> [...]
  who          [-a <anvil socket path>] [-1] [<user mask>] [<ip/bits>]
dsync-local(kishore@test.org.com): Error: read(vmail@production1.baplc.com) failed: EOF (version not received)
dsync-local(kishore@test.org.com): Error: Remote command returned error 1: ssh -lvmail production1.baplc.com doveadm dsync-server -ukishore@test.org.com

[root@production2 test.org.com]#

=================================


On Mon, 27 Jul, 2020, 11:18 pm Antonio Leding, <tech@leding.net> wrote:
Fair enough but the command snippet I posted was taken directly from your earlier post dated Mon Jul 27 23:11:06 EEST 2020 and it does say “production1”…I just double-checked as well...

FWIW…the server order in the command you just pasted is the exact reverse from the Jul 27th post I mentioned above…

Here you appear to be exec’ing the command on prod2 and trying to pull from the remote of prod 1…your earlier post was exec’ing on prod1 and pulling from the remote of prod2…






On Jul 27, 2020, at 3:10 PM, Kishore Potnuru <kishore.reachme@gmail.com> wrote:

I am executing the command from the new server. I was executing the command from production2 server only. It has the backup option, but whereas production1 doesn't have it. Though I am executing from production2, I am seeing the below error. Is it becasue production1 server doesn't have that option? Please let me know, if it is confusing.


The new version is:
====================
[root@production2 ~]# dovecot --version
2.2.36 (1f10bfa63)


The old version is : 
======================
[root@production1 ~]# dovecot --version
2.0.9


I have executed the command from the new server:



=================================
[root@production2 test.org.com]# doveadm backup -R -u kishore@test.org.com remote:vmail@production1.baplc.com
vmail@production1.baplc.com's password:
usage: doveadm [-Dv] [-f <formatter>] <command> [<args>]

  altmove      [-u <user>|-A] [-S <socket_path>] <search query>
  auth         [-a <auth socket path>] [-x <auth info>] <user> [<password>]
  config       [doveconf parameters]
  director     add|flush|map|remove|status
  dump         [-t <type>] <path>
  expunge      [-u <user>|-A] [-S <socket_path>] <search query>
  fetch        [-u <user>|-A] [-S <socket_path>] <fields> <search query>
  force-resync [-u <user>|-A] [-S <socket_path>] <mailbox>
  help         <cmd>
  import       [-u <user>|-A] [-S <socket_path>] <source mail location> <dest parent mailbox> <search query>
  kick         [-a <anvil socket path>] [-f] <user mask>[|]<ip/bits>
  log          find|reopen|test
  mailbox      create|delete|list|mutf7|rename|status|subscribe|unsubscribe
  penalty      [-a <anvil socket path>] [<ip/bits>]
  purge        [-u <user>|-A] [-S <socket_path>]
  pw           [-l] [-p plaintext] [-r rounds] [-s scheme] [-u user] [-V]
  reload
  search       [-u <user>|-A] [-S <socket_path>] <search query>
  sis          deduplicate|find
  stop
  user         [-a <userdb socket path>] [-x <auth info>] <user mask> [...]
  who          [-a <anvil socket path>] [-1] [<user mask>] [<ip/bits>]
dsync-local(kishore@test.org.com): Error: read(vmail@production1.baplc.com) failed: EOF (version not received)
dsync-local(kishore@test.org.com): Error: Remote command returned error 1: ssh -lvmail production1.baplc.com doveadm dsync-server -ukishore@test.org.com

[root@production2 test.org.com]#

=================================

On Mon, Jul 27, 2020 at 10:55 PM Antonio Leding <tech@leding.net> wrote:
Not sure if this matters but the prompt when attempting the “doveadm backup” command shows it to be a “production" server…

[root at production1 test.org.com]# doveadm backup -R -u kishore at test.org.com
remote:vmail at production2.baplc.com

...that would seem to me to indicate the old box.  If so, then per Kishore’s earlier posts, the machine where this is being attempted is running 2.0.9…not 2.2.36

Maybe nothing but thought I would mention it…




On Jul 27, 2020, at 2:50 PM, Sami Ketola <sami.ketola@dovecot.fi> wrote:



On 28. Jul 2020, at 0.10, Kishore Potnuru <kishore.reachme@gmail.com> wrote:

Yes. Please find the details of the servers below. I am executing the command from one of the new servers.

I find it very hard to believe this. Shall I look at the version control when backup was added as option to doveadm?

2.2.36 has it:

[root@ketola ~]# dovecot --version
2.2.36.8 (0c9d56b41)
[root@ketola ~]# doveadm backup
doveadm backup [-u <user>|-A] [-S <socket_path>]  [-fPRU] [-l <secs>] [-r <rawlog path>] [-m <mailbox>] [-g <mailbox_guid>] [-n <namespace> | -N] [-x <exclude>] [-s <state>] [-t <start date>] -d|<dest>


Sami