Hi,
We have some custom ROXY extensions to support cluster management activities
like replicate databases and remove database replications.
I have these below two methods to support replication removal and this works
fine when both the clusters are available.
def remove_database_replication(database = @properties["ml.content-db"])
get_admin_password
get_dr_admin_password
begin
params = {:database_name => database}
url =
%Q{http://#{@properties["ml.dr-host"]}:#{@properties["ml.app-port"]}/admin/delete-database-master.xqy}
url = url_with_parameters(url,params)
logger.info "Removing replication for #{database} from
#{@properties["ml.server"]} to #{@properties["ml.dr-host"]} - replica side "
go
url,'get',{},nil,nil,false,@properties["ml.dr-admin-user"],@properties["ml.dr-admin-password"]
rescue => error
logger.info "#{error.class} and #{error.message}"
logger.info "An error occurred when trying to remove database
replication for database #{database} from replica side. Investigate the logs
and admin screens. You may have to perform this step manually. If you are in a
DR situation this may relate to host unavailability"
end
sleep 30
begin
params = {:database_name => database}
url =
%Q{http://#{@properties["ml.server"]}:#{@properties["ml.app-port"]}/admin/delete-database-replication.xqy}
url = url_with_parameters(url,params)
logger.info "Removing replication for #{database} from
#{@properties["ml.server"]} to #{@properties["ml.dr-host"]} - master side"
r =
go(url,'get',{},nil,nil,false,@properties["ml.user"],@properties["ml.password"])
rescue RuntimeError
logger.info "An error occurred when trying to remove database
replication for database #{database} from master side. Investigate the logs and
admin screens. You may have to perform this step manually. If you are in a DR
situation this may relate to host unavailability"
end
end
def remove_database_replications
dbnames = @properties["ml.replicate-dbs"].split(",")
#Need to replicate the content database
remove_database_replication @properties["ml.content-db"]
#Need to replicate all other databases that are configured in
properties file
(dbnames).each do |dbname|
remove_database_replication dbname
end
end
But I when shutdown the primary and try to execute these, I expect that DR side
invocation will be successful and Primary side call fails as it is not
reachable. That seems to work fine for 1st database and for the subsequent
databases, the http call is not at all happening though I see a log message
before the go method invocation.
Enter password for admin on primaryhostname : (Production)
Enter password for admin on drhostname : (DR)
Removing replication for engineering-lab-content from primaryhostname to
drhostname - replica side
Request made to http://drhostname:8041 using user_name admin
Removing replication for engineering-lab-content from primaryhostname to
drhostname - master side
Request made to http://primaryhostname:8041 using user_name admin
Roxy::Http : request failure count: 1, exception: #<Errno::ECONNREFUSED:
Connection refused - connect(2)>
Roxy::Http : request failure count: 2, exception: #<Errno::ECONNREFUSED:
Connection refused - connect(2)>
Roxy::Http : request failure count: 3, exception: #<Errno::ECONNREFUSED:
Connection refused - connect(2)>
Roxy::Http : request failure count: 4, exception: #<Errno::ECONNREFUSED:
Connection refused - connect(2)>
Roxy::Http : re-raising same error: primaryhostname temporarily unavailable:
(Connection refused - connect(2)) -- error count: 4, error age: 0
An error occurred when trying to remove database replication for database
engineering-lab-content from master side. Investigate the logs and admin
screens. You may have to perform this step manually. If you are in a DR
situation this may relate to host unavailability
Removing replication for Security from primaryhostname to drhostname - replica
side
Request made to http://drhostname:8041 using user_name admin
Roxy::Http : re-raising same error: primaryhostname temporarily unavailable:
(Connection refused - connect(2)) -- error count: 4, error age: 0
RuntimeError and primaryhostname temporarily unavailable: (Connection refused -
connect(2))
An error occurred when trying to remove database replication for database
Security from replica side. Investigate the logs and admin screens. You may
have to perform this step manually. If you are in a DR situation this may
relate to host unavailability
Removing replication for Security from primaryhostname to drhostname - master
side
Request made to http://primaryhostname:8041 using user_name admin
Roxy::Http : request failure count: 5, exception: #<Errno::ECONNREFUSED:
Connection refused - connect(2)>
Roxy::Http : re-raising same error: primaryhostname temporarily unavailable:
(Connection refused - connect(2)) -- error count: 5, error age: 0
An error occurred when trying to remove database replication for database
Security from master side. Investigate the logs and admin screens. You may have
to perform this step manually. If you are in a DR situation this may relate to
host unavailability
Removing replication for Triggers from primaryhostname to drhostname - replica
side
Request made to http://drhostname:8041 using user_name admin
Roxy::Http : re-raising same error: primaryhostname temporarily unavailable:
(Connection refused - connect(2)) -- error count: 5, error age: 0
RuntimeError and primaryhostname temporarily unavailable: (Connection refused -
connect(2))
An error occurred when trying to remove database replication for database
Triggers from replica side. Investigate the logs and admin screens. You may
have to perform this step manually. If you are in a DR situation this may
relate to host unavailability
Removing replication for Triggers from primaryhostname to drhostname - master
side
Request made to http://primaryhostname:8041 using user_name admin
Roxy::Http : request failure count: 6, exception: #<Errno::ECONNREFUSED:
Connection refused - connect(2)>
Roxy::Http : re-raising same error: primaryhostname temporarily unavailable:
(Connection refused - connect(2)) -- error count: 6, error age: 0
An error occurred when trying to remove database replication for database
Triggers from master side. Investigate the logs and admin screens. You may have
to perform this step manually. If you are in a DR situation this may relate to
host unavailability
Green color highlights first call for content database on the replica side that
is successful. Yellow color highlights the call that are expected to be
successful for DR but never seen reaching the DR MarkLogic server side and ROXY
just spits the old error message of connection error for master host.
I am new to ruby and appreciate your help in solving this. I have looked into
RoxyHttp.rb file request method a little and It looks like it is coded that
way. I would like to get around this and treat new requests differently.
Thanks.
Regards,
Sudheer Yalaverthi
NoSQL Engineer
---
This communication may contain confidential and/or privileged information. If
you are not the intended recipient (or have received this communication in
error) please notify the sender immediately and destroy this communication. Any
unauthorized copying, disclosure or distribution of the material in this
communication is strictly forbidden.
Deutsche Bank does not render legal or tax advice, and the information
contained in this communication should not be regarded as such.
_______________________________________________
General mailing list
[email protected]
Manage your subscription at:
http://developer.marklogic.com/mailman/listinfo/general
