[pgpool-general: 7017] Re: Pgpool-II-4.1.1 and Ubuntu 20.04

Wolf Schwurack wolf at uen.org
Thu May 14 01:33:50 JST 2020


I installed pgpool-II-4.0.8 using pgpool.conf from pgpool-II-4.1.1 watchdog failed to start. But when I used pgpool.conf from pgpool-II-4.0.8 watchdog started on a single node

2020-05-13 10:28:50: pid 11671: LOG:  pgpool-II successfully started. version 4.0.8 (torokiboshi)
2020-05-13 10:28:50: pid 11671: LOG:  node status[0]: 1
2020-05-13 10:28:50: pid 11671: LOG:  node status[1]: 2
2020-05-13 10:28:51: pid 11685: LOG:  creating socket for sending heartbeat
2020-05-13 10:28:51: pid 11685: DETAIL:  bind send socket to device: ens160
2020-05-13 10:28:51: pid 11685: LOG:  set SO_REUSEPORT option to the socket
2020-05-13 10:28:51: pid 11685: LOG:  creating socket for sending heartbeat
2020-05-13 10:28:51: pid 11685: DETAIL:  set SO_REUSEPORT
2020-05-13 10:28:51: pid 11684: LOG:  createing watchdog heartbeat receive socket.
2020-05-13 10:28:51: pid 11684: DETAIL:  bind receive socket to device: "ens160"
2020-05-13 10:28:51: pid 11684: LOG:  set SO_REUSEPORT option to the socket
2020-05-13 10:28:51: pid 11684: LOG:  creating watchdog heartbeat receive socket.
2020-05-13 10:28:51: pid 11684: DETAIL:  set SO_REUSEPORT
2020-05-13 10:28:57: pid 11681: LOG:  successfully acquired the delegate IP:"10.11.0.204"
2020-05-13 10:28:57: pid 11681: DETAIL:  'if_up_cmd' returned with success
2020-05-13 10:28:57: pid 11673: LOG:  watchdog escalation process with pid: 11681 exit with SUCCESS.


Wolfgang Schwurack
Database/System Administrator
Utah Education Network
801-587-9444
wolf at uen.org
 


On 5/13/20, 9:39 AM, "pgpool-general-bounces at pgpool.net on behalf of Wolf Schwurack" <pgpool-general-bounces at pgpool.net on behalf of wolf at uen.org> wrote:

    Still testing this I decided to start pgpool on the second node which started watchdog. I know you told me before about having an even number a nodes to set "enable_consensus_with_half_votes = on" Which is how this is set in pgpool.conf. But I have 3 nodes setup which is uneven number of nodes. 
    
    It seems that the parameter "enable_consensus_with_half_votes" is not working. I can set it to 'on or off' The delegate IP is only successfully acquired if the second node is running pgpool 
    
    Here's the log output once I start pgpool on the second node. 
    
    2020-05-13 09:11:08: pid 4103: LOG:  new watchdog node connection is received from "10.11.0.206:17581"
    2020-05-13 09:11:08: pid 4103: LOG:  new node joined the cluster hostname:"pgtest-02" port:9000 pgpool_port:9999
    2020-05-13 09:11:08: pid 4103: DETAIL:  Pgpool-II version:"4.1.1" watchdog messaging version: 1.1
    2020-05-13 09:11:08: pid 4103: LOG:  new outbound connection to pgtest-02:9000 
    2020-05-13 09:11:14: pid 4103: LOG:  adding watchdog node "pgtest-02:9999 Linux pgtest-02" to the standby list
    2020-05-13 09:11:14: pid 4103: LOG:  quorum found
    2020-05-13 09:11:14: pid 4103: DETAIL:  starting escalation process
    2020-05-13 09:11:14: pid 4103: LOG:  escalation process started with PID:4231
    2020-05-13 09:11:14: pid 4099: LOG:  Pgpool-II parent process received watchdog quorum change signal from watchdog
    2020-05-13 09:11:14: pid 4103: LOG:  new IPC connection received
    2020-05-13 09:11:14: pid 4099: LOG:  watchdog cluster now holds the quorum
    2020-05-13 09:11:14: pid 4099: DETAIL:  updating the state of quarantine backend nodes
    2020-05-13 09:11:14: pid 4103: LOG:  new IPC connection received
    2020-05-13 09:11:14: pid 4231: LOG:  watchdog: escalation started
    2020-05-13 09:11:21: pid 4231: LOG:  successfully acquired the delegate IP:"10.11.0.204"
    2020-05-13 09:11:21: pid 4231: DETAIL:  'if_up_cmd' returned with success
    2020-05-13 09:11:21: pid 4103: LOG:  watchdog escalation process with pid: 4231 exit with SUCCESS.
    
    Then I start pgpool on the third node
    2020-05-13 09:26:52: pid 4906: LOG:  new watchdog node connection is received from "10.11.0.207:47278"
    2020-05-13 09:26:52: pid 4906: LOG:  new node joined the cluster hostname:"pgtest-03" port:9000 pgpool_port:9999
    2020-05-13 09:26:52: pid 4906: DETAIL:  Pgpool-II version:"4.1.1" watchdog messaging version: 1.1
    2020-05-13 09:26:52: pid 4906: LOG:  new outbound connection to pgtest-03:9000 
    2020-05-13 09:26:58: pid 4906: LOG:  adding watchdog node "pgtest-03:9999 Linux pgtest-03" to the standby list
     
    Now when I stop pgpool on third node
    2020-05-13 09:34:19: pid 4906: LOG:  client socket of pgtest-03:9999 Linux pgtest-03 is closed
    2020-05-13 09:34:19: pid 4906: LOG:  remote node "pgtest-03:9999 Linux pgtest-03" is shutting down
    2020-05-13 09:34:19: pid 4906: LOG:  removing watchdog node "pgtest-03:9999 Linux pgtest-03" from the standby list
    
    Watchdog is still running
    root at pgtest-01:~# ip a
    1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
        link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
        inet 127.0.0.1/8 scope host lo
           valid_lft forever preferred_lft forever
    2: ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
        link/ether 00:50:56:b3:d3:58 brd ff:ff:ff:ff:ff:ff
        inet 10.11.0.205/24 brd 10.11.0.255 scope global ens160
           valid_lft forever preferred_lft forever
        inet 10.11.0.204/24 brd 10.11.0.255 scope global secondary ens160:1
           valid_lft forever preferred_lft forever
    
    Now when I stop pgpool on the second node
    2020-05-13 09:36:01: pid 4906: LOG:  We have lost the quorum
    2020-05-13 09:36:01: pid 4902: LOG:  Pgpool-II parent process received watchdog quorum change signal from watchdog
    2020-05-13 09:36:01: pid 4906: LOG:  new IPC connection received
    2020-05-13 09:36:01: pid 5504: LOG:  watchdog: de-escalation started
    2020-05-13 09:36:01: pid 5504: LOG:  successfully released the delegate IP:"10.11.0.204"
    2020-05-13 09:36:01: pid 5504: DETAIL:  'if_down_cmd' returned with success
    2020-05-13 09:36:01: pid 4906: LOG:  watchdog de-escalation process with pid: 5504 exit with SUCCESS.
    
    _______________________________________________
    pgpool-general mailing list
    pgpool-general at pgpool.net
    http://www.pgpool.net/mailman/listinfo/pgpool-general
    



More information about the pgpool-general mailing list