[pgpool-general: 7368] Re: User "pgpool" does not have a valid SCRAM

[Luís Alves]

[1-1] 2020-12-17 17:23:59: pid 9732: LOG:  reading status file: 0 th
backend is set to down status  <--- I think this is the problem

Now I just need to find the file location.

Regards,
LA

On Thu, Dec 17, 2020 at 4:19 PM Luís Alves <luisalves00 at gmail.com> wrote:

> is there any file that store the index of the primary node?
>
> Regards,
> LA
>
> On Thu, Dec 17, 2020 at 3:56 PM Luís Alves <luisalves00 at gmail.com> wrote:
>
>> The problem is not the recovery node index. Now I have:
>>
>> pcp_recovery_node -w -h 10.230.4.148 -p 9898 -U pgpool -n 1
>>
>> (I'm going to try to delay this action on ansible, as pgpool might not
>> have enough time to find the primary node)
>>
>> Then:
>>
>> -bash-4.2$ psql -h 10.230.4.148 -p 9999 -U pgpool postgres -c "show
>> pool_nodes"
>> psql: FATAL:  failed to create a backend 1 connection
>> DETAIL:  not executing failover because failover_on_backend_error is off
>>
>> checking node 0:
>>
>> -bash-4.2$ pcp_node_info -w -h 10.3.84.4 -p 9898 -U pgpool -n 0
>> ERROR: connection to host "10.3.84.4" failed with error "Connection
>> refused"
>>
>> But the server is running:
>>
>> qs-auth-01 ~]# service postgresql-11 status
>> Redirecting to /bin/systemctl status postgresql-11.service
>> ● postgresql-11.service - PostgreSQL 11 database server
>>    Loaded: loaded (/usr/lib/systemd/system/postgresql-11.service;
>> disabled; vendor preset: disabled)
>>   Drop-In: /etc/systemd/system/postgresql-11.service.d
>>            └─override.conf
>>    Active: active (running) since Thu 2020-12-17 16:43:48 CET; 8min ago
>>
>> but pgpool is only looking into qs-auth-02 (which is node index 1)
>>
>> Any ideas?
>>
>> Regards,
>> LA
>>
>>
>> On Thu, Dec 17, 2020 at 1:41 PM Luís Alves <luisalves00 at gmail.com> wrote:
>>
>>> I'm doing the initial installation using Ansible.
>>> I think the problem was that on the initial installation where node 0
>>> (qs-auth-01.dc.internal) should be the primary I was executing:
>>>
>>> pcp_recovery_node -w -h {{ delegate_virtual_ip }} -p 9898 -U pgpool -n 0
>>>
>>> instead of
>>>
>>> pcp_recovery_node -w -h {{ delegate_virtual_ip }} -p 9898 -U pgpool -n 1
>>>
>>> because my node index calculation had a bug.
>>> I'll check it out during this afternoon and I'll update the thread if
>>> the problem persists.
>>>
>>> Regards,
>>> LA
>>>
>>>
>>>
>>> On Thu, Dec 17, 2020 at 12:21 PM Tatsuo Ishii <ishii at sraoss.co.jp>
>>> wrote:
>>>
>>>> > Now they are correct and it can connect. Nevertheless, I'm having
>>>> > other issues. Here is the Backend Connection Settings
>>>> >
>>>> >
>>>> > # - Backend Connection Settings -
>>>> > backend_hostname0 = 'qs-auth-01.dc.internal'
>>>> > backend_port0 = 5432
>>>> > backend_weight0 = 1
>>>> > backend_data_directory0 = '/export/pgsql/data'
>>>> > backend_flag0 = 'ALLOW_TO_FAILOVER'
>>>> > backend_application_name0 = 'qs-auth-01.dc.internal'
>>>> > backend_hostname1 = 'qs-auth-02.dc. internal '
>>>> > backend_port1 = 5432
>>>> > backend_weight1 = 1
>>>> > backend_data_directory1 = '/export/pgsql/data'
>>>> > backend_flag1 = 'ALLOW_TO_FAILOVER'
>>>> > backend_application_name1 = 'qs-auth-02.dc.internal'
>>>> >
>>>> >
>>>> > node qs-auth-01.dc.internal should be the primary and has postgres
>>>> running
>>>> > there, but pgpool seems to ignore it.
>>>> > node qs-auth-02.dc.internal is shutdown as I want to apply
>>>> replication from
>>>> > node 1.
>>>> >
>>>> > Here is the log:
>>>> >
>>>> > Dec 17 09:18:32 qs-auth-02 pgpool[20471]: [915-1] 2020-12-17
>>>> 09:18:32: pid
>>>> > 20471: ERROR:  failed to make persistent db connection
>>>> > Dec 17 09:18:32 qs-auth-02 pgpool[20471]: [915-2] 2020-12-17
>>>> 09:18:32: pid
>>>> > 20471: DETAIL:  connection to host:"qs-auth-02.dc.internal:5432"
>>>> failed
>>>> > Dec 17 09:18:32 qs-auth-02 pgpool[20471]: [916-1] 2020-12-17
>>>> 09:18:32: pid
>>>> > 20471: LOG:  health check retrying on DB node: 1 (round:2)
>>>> > Dec 17 09:18:33 qs-auth-02 pgpool[20471]: [917-1] 2020-12-17
>>>> 09:18:33: pid
>>>> > 20471: LOG:  failed to connect to PostgreSQL server on
>>>> > "qs-auth-02.dc.internal:5432", getsockopt() detected error "Connection
>>>> > refused"
>>>> > Dec 17 09:18:33 qs-auth-02 pgpool[20471]: [918-1] 2020-12-17
>>>> 09:18:33: pid
>>>> > 20471: ERROR:  failed to make persistent db connection
>>>> > Dec 17 09:18:33 qs-auth-02 pgpool[20471]: [918-2] 2020-12-17
>>>> 09:18:33: pid
>>>> > 20471: DETAIL:  connection to host:"qs-auth-02.dc.internal:5432"
>>>> failed
>>>> > Dec 17 09:18:33 qs-auth-02 pgpool[20471]: [919-1] 2020-12-17
>>>> 09:18:33: pid
>>>> > 20471: LOG:  health check retrying on DB node: 1 (round:3)
>>>> > Dec 17 09:18:34 qs-auth-02 pgpool[20471]: [920-1] 2020-12-17
>>>> 09:18:34: pid
>>>> > 20471: LOG:  failed to connect to PostgreSQL server on
>>>> > "qs-auth-02.dc.internal:5432", getsockopt() detected error "Connection
>>>> > refused"
>>>> >
>>>> > I see nothing in the node qs-auth-01.dc.internal postgres logs. Is
>>>> there
>>>> > another place where pgpool looks to find postgres servers?
>>>>
>>>> Probably pgpool recognizes backend_hostname0 (qs-auth-01.dc.internal)
>>>> as in "down" or "unused" status.  You can verify it by executing "show
>>>> pool_nodes" from psql command connecting to pgpool.
>>>>
>>>> Best regards,
>>>> --
>>>> Tatsuo Ishii
>>>> SRA OSS, Inc. Japan
>>>> English: http://www.sraoss.co.jp/index_en.php
>>>> Japanese:http://www.sraoss.co.jp
>>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20201217/9713ca94/attachment-0001.htm>