Tag Info

Hot answers tagged

10

There's no such thing as zero downtime. Set a realistic goal (say, five nines uptime), and plan around that. If you beat your target, great, but promising a system will never, ever go down can trap you in a state where it's impossible to make major architectural upgrades necessary to continue maintaining the system at reasonable expense.


9

Thoughts: If you are asking these question on a public forum, hire experts to do it for you Consider Postgres and SQL Server which will scale to this volume too Do you need ACID? No = consider NoSQL Design and hardware matter more than the platform Don't virtualise or cut other hardware corners What is your RPO/RTO? Maintenance window? a.k.a are you really ...


3

It should be safe enough to simply remove the Secondary Namenode role, and then add it back again onto a different node on the cluster. In the intermediate period, you may see a warning from Cloudera Manager that the role doesn't exist (which may cause long term issues with the namenode), but not having a Secondary doesn't put your data any more at risk.


3

It is required during the installation and upgrade procedure when using the wizard in the Cloudera Manager. During normal cluster operation it is not needed. All communication is done via the locally running agents then. In fact Cloudera Manager does not store the credentials, after the installation / upgrade it will delete it. When you upgrade your ...


3

Looks like I had mistakenly set fs.default.name to 9000. Changed it to 8020, now everything works fine. Thanks to mgorven for the help!


2

The main consideration for zero (or close to) downtime is the amount of update activity. Updates (and deletes) conflict in ways that inserts don't. Level 1. Database is pretty much totally read only (eg used for a Content Management System). This is the easiest to replicate. Level 2. Only Inserts on a single node which get distributed to other 'read-only' ...


2

I think my replication went bad. I seems that .oldlogs is where the Write-Ahead-Logs (WALS) go according to this safari article. They should be cleaned up but were not for some reason. I used the following to clean it up: HADOOP_USER_NAME=hdfs hadoop fs -rm -skipTrash /hbase/.oldlogs/* Since I noticed this while working on building a replacement cluster ...


2

The job of the 2NN is to read the changes on the HDFS filesystem and add them to the fsimage. This decreases NN startup time as during startup the NN reads the fsimage file and then applies all the interim log changes on top of it. The naming is somewhat unfortunate as it is indeed not a standby/backup NN but only a utility to increase NN performance. In ...


1

HBase is crash-safe and .logs is the location of WALs (hlogs) that are needed for crash recovery. Once the memory of regionservers is flushed to hfiles, WALs are no longer needed for crash recovery and they are moved to .oldlogs. Old logs are usually used for cluster-to-cluster replication. .oldlogs have a configurable retention period, for example 3 days. ...


1

tcp_v4_rcv [0] calls sk_add_backlog and if it fails it increments TCPBacklogDrop 2014 } else if (unlikely(sk_add_backlog(sk, skb, 2015 sk->sk_rcvbuf + sk->sk_sndbuf))) { 2016 bh_unlock_sock(sk); 2017 NET_INC_STATS_BH(net, LINUX_MIB_TCPBACKLOGDROP); 2018 ...


1

Follow these two steps: 1) Change the replication factor from hadoop file system. Make sure to login to the user for which you are getting health issue of under-replicated blocks: su - hdfs hadoop fs -setrep -R 1 / or su - oozie hadoop fs -setrep -R 1 / etc... 2) Change Warning and Critical value for "Under-replicated Block Monitoring Thresholds" ...


1

According to this documentation the /usr/lib/hbase/conf/hbase-site.xml file needs to be configured as follows: <configuration> <property> <name>hbase.cluster.distributed</name> <value>true</value> </property> </configuration> to avoid HBase manages its own ZooKeeper. After configuring this, the ...


1

The answer is hadoop-0.20.205.0 but you also need to replace jar files in ./lib/ of HBase with the ones of your working HDFS installation. For me it was: rm /opt/hbase-0.90.4/lib/hadoop-core-0.20-append-r1056497.jar cp /opt/hadoop-0.20.205.0/hadoop-core-0.20.205.0.jar /opt/hbase-0.90.4/lib/ cp /opt/hadoop-0.20.205.0/commons-configuration-1.6.jar ...


1

People usually use the Hadoop framework to process data coming out of scribe. You can use PHP to script map/reduce steps... However, unless you have a lot of data, this is probably going to be a lot of work for a small benefit: http://hadoop.apache.org/ Good luck with that !


1

First check "hbase-hduser-master-ubuntu.log" file from logs folder of your hbase installed path if you are getting following error 2014-09-07 20:22:24,505 ERROR [main] master.HMasterCommandLine: Master exiting java.lang.RuntimeException: HMaster Aborted at org.apache.hadoop.hbase.master.HMasterCommandLine.startMaster(HMasterCommandLine.java:194) at ...


1

From the comments - this problem was caused because zookeeper was not running.


1

I'd say it is possible to achieve zero downtime. GoldenGate tries to provide this solution with bi-directional replication. You'll still need conflict resolution for an active-active configuration, and yes it can become a problem, but it's a pretty good solution. For master/slave ChronicDB can do live updates accounting for replication without ...


1

The easiest way to achieve an active/active configuration in Oracle is by using Oracle RAC (Real Application Cluster). RAC Documentation can be found here. RAC can also be combined with other Oracle High Availability tools like Data Guard or Streams. HA Documentation is available here. Keep in mind that some maintenance operation can require you to shut ...


1

This isn't necessarily a serverfault question. Being able to run two active databases in parallel will depend on your application code. The trick is you must design you code so there will never be a conflict with records simultaneously changed in both locations. A few design ideas: Use SYS_GUID to create all your primary keys instead of a sequence which ...



Only top voted, non community-wiki answers of a minimum length are eligible