we are using two Memcached servers clusterted (Master , Master ) and used repcached on top of it for replication purpose , for our heavy traffic website
For all logged in users , I am doing a atomic Memcached cas Operation as shown below .
long casVal = resp.getCas();
MemcacheClinet.cas(TOTAL_USERS, 0,value, casVal);
After doing some research using Memcache Stats Commands on both the servers using get TOTAL_USERS , i found that this key has got a different value on each server and its not being replicated properly and its behaving abnormally
Please let me know if is it possible to get this key "TOTAL_USERS" replicated properly on both the servers ??
Or is there any other way where i can manually append the Data to this String ??
I am uisng XMemcached 1.4.5 version with repcached 2.2 installed on to it .
Please share your ideas .
Related
I'm getting this error while using Monstache:
Unable to create Elasticsearch client: health check timeout: no Elasticsearch node available
I applied these lines to Monstache configuration:
elasticsearch-validate-pem-file = false
elasticsearch-healthcheck-timeout-startup = 200
elasticsearch-healthcheck-timeout = 200
However, I still encounter the mentioned error. When I searched about it, I found that the problem is due to sniffing in elasticsearch client. But I don't know where and how exactly I must change it?
I should denote that I studied this tutorial for this problem, but I'm still full of ambiguities.
The problem has been solved when I installed Monstache on the same local server on which the ELK stack was installed. Also, the MongoDB database on the remote server has been changed to a single node replica set to be able to connect to Monstache.
let's try to use
elastic.SetSniff(false)
In the PostgreSQL documentation https://www.postgresql.org/docs/10/libpq-connect.html, it has been said that multiple hosts can be specified in a single connection string such that all the hosts will be tried in order one after the other until one of the server gets succeeds.
But when i tried to implement the same setting in the tag present in my ASP.net web.config file, it is throwing error as no such host name. I am using NpgSQL provider in order to connect to PostgreSQL database.
I need to add multiple server names in the connection string such that if the server#1 fails then it should try for the next server server#2 immediately provided in the order until it succeeds
Can you please suggest on how multiple hosts can be provided in the connection string?
The Npgsql driver does not currently support this functionality. The issue tracking this is https://github.com/npgsql/npgsql/issues/732, I'm still hoping we can get this into the next release but there's a lot going on.
Load balancing and failover is avaialble in Npgsql version 6. At the time of writing v.6 is in preview.
Simple failover example (server2 is only used if a connection could not be established to server1):
Host=server1,server2;Username=test;Password=test
Example with load balancing (round robin I guess):
Host=server1,server2,server3,server4,server5;Username=test;Password=test;Load
Balance Hosts=true;Target Session Attributes=prefer-standby
https://www.npgsql.org/doc/failover-and-load-balancing.html
I have react frontend and spring boot backend with mongodb behind.
I have issues with setting 2 parameters in the spring boot service.
First is address of the mongodb which is now set as localhost:27017 in the application.properties
It works at localhost but since I plan to scale out using kubernetes and docker images i would like to know how to define
It and where for the case in which I have mongo1 mongo2 and mongo3 database hosts and would like to pass all 3 URIs ?
Second issues is more tricky! React frontend doesnt work in chrome until I put allow cross origin anotation over my spring rest endpoint . I used hardcoded localhost:3000 here but when I scale it out using kubernetes this wont work if it gets data from another host in the cluster.What to do here?
To answer your first question, you can configure multi data sources, see here documentation how you can configure more than one data sources (80.2 Configure Two DataSources.
For second question you can simply wildcard CORS URL or if you know all of your front end server urls which are load balanced you can pass as list of cors url.
– * – means that all origins are allowed.
– If undefined, all origins are allowed.
RECOMMENDATION
Run your react via yarn to deploy on Apache or ngnix. Once you seted up your domain or sub domain for front end, load balanced your front end so not required to run your front end on ports..
I did move the databases from our ServiceBus test enviroment.
I started by leaving the farm with the single node, then I moved the databases.
After rejoining the farm I see that GatewayDBConnectionString is till pointing to the old one.
I can't find any valid PowerShell command to reconfigure the value in question.
Anyone know how to fix this?
Thank you in advance.
To answer this I will need you to understand this a bit more - and hence giving a high-level overview of Service Bus 1.1 Server farm configuration:
Service Bus Server 1.1 is a platform where users can create highly-durable distributed Pub-Sub (messaging Queues/Topics) entities. In simple words - the main job of this is to translate the Compute (your VMs) and Data (your MsgContainer databases) into messaging functionality Durable Queues and Topics. So, in short - the configuration wizard or the Powershell cmdlets used to configure ServiceBus 1.1 Server will try to take the VMs and Databases from you.
The Db SBManagementDB is considered to be the authoritative source of truth for any Farm level configuration -> like Nodes that are part of the Farm (Store.Nodes), Ports opened on each of the nodes, Gateway database connection string (Cluster Config) etc. Also pl. note that - as per the Windows Server product guidelines - any information that has to be securely persisted will be encrypted - so as the Gateway DB connection String.
a) when you did New-SBFarm (with a Gateway DB connection string) - you have essentially communicated to SBMgmtDB - the Gateway DB Server, database name etc.
b) when you do Add-SBHost - again you have communicated to SBMgmtDb that you want to add one Node to this Farm
Gateway db connection string is the one place for Truth for all Gateway Services to find any run-time info -> like Container Databases, entity to container mapping etc.
again, when you do New-SBMessageContainer PSCmdlet --> you communicated to SBGatewayDB that you are adding one db
Now, with this background - lets see how the action you did above will take into effect:
- When you moved all the Databases to a different Server - you changed the Gateway Database connection string - But the Gateway connection string you had communicated to the SBManagementDB (using the New-SBFarm cmdlet) was pointing to the Old Server.
- When you removed the Node from the Farm and again Joined back - you removed one node from the configuration and re-added it - no affect :)
The ANSWER
Use Restore-SBFarm PS Cmdlet to communicate to the SBManagementDB that you changed the GW db
and then Use Restore-SBMessageContainer PS Cmdlet to communicate to Gateway DB that you changed the Container databases.
Now, add the Nodes back to this restored farm.
HTH!
Sree
I have successfully setup ndbcluster version 7.1.26.
This contains 2 data nodes[NDBD], 2 mysql [MYSQLD] nodes and one management [MGMD] node.
Replication works successfully.
My Web application is deployed in JBoss-5.0.1 and using JNDI for connection resources which are specified in application specific ds.xml file in load balanced url forms e.g. jbdc:mysql:loadbalance:host1:port1,host2:port2/databaseName.
host1 : refers to first mysqld node and port1 refers the port it is running on.
host2 : refers to second mysqld node and port2 refers the port it is running on.
When both of the [MySQLD] nodes are up and running everything works fine and cluster responds well, replicates data, and data retrieval operations also work properly.
But issues are raised when any of the [MySQLD] nodes goes down. Data gets inserted/updated/replicated but the application is unable to retrieve data from cluster and web page remains busy working which means busy retrieving data. As soon as the node which was down goes up it responds properly and application goes forward and shows up data retrieved from cluster.
At JBoss 5.0.1 startup it showed up a NullPointerException in class LoadBalancingConnectionProxy.invoke(LoadBalancingConnectionProxy.java:439). Tell me if the above Exception plays any role in the above explained issues.
If anyone had faced issues like above and if has any solution regarding the issues please let me know.
Thanks and regards.
I have resolved the issue as it was a bug in the connectorJ's version.
As The project I am working on was already using both the buggy jar mysql-connector-java-5.0.8.jar and the jar version in which the issue is already resolved i.e. mysql-connector-java-5.1.13-bin.jar.
After all the search when I removed the jar mysql-connector-java-5.0.8.jar my issues got resolved.
All that was problematic was that the ConnectorJ/Driver was getting referred from the buggy jar.
The bug id and url which refers to this issue is:
http://bugs.mysql.com/bug.php?id=31053
.
Thanks for considerations.
Are you using different userids and passwords for each of the hosts(host1, host2) specified in the tag ? (Either directly or using tag) ?