restart cassandra after instances stop/start
Hi, I followed the instructions here http://nubetech.co/setting-up-cassandra-clusters-in-the-cloud-with-whirr to set up an ec2 cluster on AWS. Everything works great until I stop my instances (to save money) and restart them because the instances get new IPs. What changes do I need to make in order to get my cassandra cluster up and running again? Thanks. Regards, -k
Re: restart cassandra after instances stop/start
I see.. What if I only need to resume the cluster and don't worry about any data? Each time I use whirr to set up a new clusters, new EBS volumes are also created and I got charged for the volumes at full month rate even though I only use them for a few hours.Thanks. -k On Sun, Jun 5, 2011 at 5:43 PM, Andrei Savu wrote: > Hi, > > There is no easy way to suspend a cluster started with Whirr without > loosing all the data stored on the nodes. You need a custom tool to > export and save data on shutdown (maybe to S3) and load it back when > you start the cluster again. > > We are tracking progress on this issue here: > https://issues.apache.org/jira/browse/WHIRR-309 > > Cheers, > > -- Andrei Savu / andreisavu.ro > > On Mon, Jun 6, 2011 at 12:34 AM, Khanh Nguyen > wrote: >> Hi, >> >> I followed the instructions here >> http://nubetech.co/setting-up-cassandra-clusters-in-the-cloud-with-whirr >> to set up an ec2 cluster on AWS. Everything works great until I stop >> my instances (to save money) and restart them because the instances >> get new IPs. What changes do I need to make in order to get my >> cassandra cluster up and running again? Thanks. >> >> Regards, >> >> -k >> >
Re: restart cassandra after instances stop/start
I add another line to my cassandra.properties whirr.image-id= with being several AMIs from http://cloud.ubuntu.ami/ but I always got an exception ..., description=099720109477/ebs/ubuntu-images/ubuntu-lucid-10.04-i386-server-20100923], description=099720109477/ebs/ubuntu-images/ubuntu-lucid-10.04-i386-server-20100923, version=20100923, location=[id=us-east-1, scope=REGION, description=us-east-1, parent=aws-ec2, iso3166Codes=[US-VA], metadata={}], loginUser=ubuntu, userMetadata={owner=099720109477, rootDeviceType=ebs}]] at org.jclouds.compute.domain.internal.TemplateBuilderImpl.build(TemplateBuilderImpl.java:551) at org.apache.whirr.cluster.actions.BootstrapClusterAction.buildTemplate(BootstrapClusterAction.java:167) at org.apache.whirr.cluster.actions.BootstrapClusterAction.doAction(BootstrapClusterAction.java:117) at org.apache.whirr.cluster.actions.ScriptBasedClusterAction.execute(ScriptBasedClusterAction.java:74) at org.apache.whirr.service.Service.launchCluster(Service.java:76) at org.apache.whirr.cli.command.LaunchClusterCommand.run(LaunchClusterCommand.java:61) at org.apache.whirr.cli.Main.run(Main.java:67) at org.apache.whirr.cli.Main.main(Main.java:104) What am I missing? Thanks. -k On Mon, Jun 6, 2011 at 2:20 AM, Andrei Savu wrote: > To avoid that you just need to use a different AMI. Try the ones > packaged by Canonical: > http://cloud.ubuntu.com/ami/ > > I guess we should improve the automatic defaults to avoid using images > with EBS persistence. I will open an issue for that. > > Cheers, > > -- Andrei > > On Mon, Jun 6, 2011 at 7:14 AM, Khanh Nguyen wrote: >> I see.. >> >> What if I only need to resume the cluster and don't worry about any >> data? Each time I use whirr to set up a new clusters, new EBS volumes >> are also created and I got charged for the volumes at full month rate >> even though I only use them for a few hours.Thanks. >> >> -k >> >> On Sun, Jun 5, 2011 at 5:43 PM, Andrei Savu wrote: >>> Hi, >>> >>> There is no easy way to suspend a cluster started with Whirr without >>> loosing all the data stored on the nodes. You need a custom tool to >>> export and save data on shutdown (maybe to S3) and load it back when >>> you start the cluster again. >>> >>> We are tracking progress on this issue here: >>> https://issues.apache.org/jira/browse/WHIRR-309 >>> >>> Cheers, >>> >>> -- Andrei Savu / andreisavu.ro >>> >>> On Mon, Jun 6, 2011 at 12:34 AM, Khanh Nguyen >>> wrote: >>>> Hi, >>>> >>>> I followed the instructions here >>>> http://nubetech.co/setting-up-cassandra-clusters-in-the-cloud-with-whirr >>>> to set up an ec2 cluster on AWS. Everything works great until I stop >>>> my instances (to save money) and restart them because the instances >>>> get new IPs. What changes do I need to make in order to get my >>>> cassandra cluster up and running again? Thanks. >>>> >>>> Regards, >>>> >>>> -k >>>> >>> >> >
Re: restart cassandra after instances stop/start
Oh, nevermind. I figured it out. It should be whirr.image-id = / On Mon, Jun 6, 2011 at 2:48 PM, Khanh Nguyen wrote: > I add another line to my cassandra.properties > > whirr.image-id= > > with being several AMIs from http://cloud.ubuntu.ami/ but I > always got an exception > > ..., > description=099720109477/ebs/ubuntu-images/ubuntu-lucid-10.04-i386-server-20100923], > description=099720109477/ebs/ubuntu-images/ubuntu-lucid-10.04-i386-server-20100923, > version=20100923, location=[id=us-east-1, scope=REGION, > description=us-east-1, parent=aws-ec2, iso3166Codes=[US-VA], > metadata={}], loginUser=ubuntu, userMetadata={owner=099720109477, > rootDeviceType=ebs}]] > at > org.jclouds.compute.domain.internal.TemplateBuilderImpl.build(TemplateBuilderImpl.java:551) > at > org.apache.whirr.cluster.actions.BootstrapClusterAction.buildTemplate(BootstrapClusterAction.java:167) > at > org.apache.whirr.cluster.actions.BootstrapClusterAction.doAction(BootstrapClusterAction.java:117) > at > org.apache.whirr.cluster.actions.ScriptBasedClusterAction.execute(ScriptBasedClusterAction.java:74) > at org.apache.whirr.service.Service.launchCluster(Service.java:76) > at > org.apache.whirr.cli.command.LaunchClusterCommand.run(LaunchClusterCommand.java:61) > at org.apache.whirr.cli.Main.run(Main.java:67) > at org.apache.whirr.cli.Main.main(Main.java:104) > > What am I missing? Thanks. > > -k > > > > On Mon, Jun 6, 2011 at 2:20 AM, Andrei Savu wrote: >> To avoid that you just need to use a different AMI. Try the ones >> packaged by Canonical: >> http://cloud.ubuntu.com/ami/ >> >> I guess we should improve the automatic defaults to avoid using images >> with EBS persistence. I will open an issue for that. >> >> Cheers, >> >> -- Andrei >> >> On Mon, Jun 6, 2011 at 7:14 AM, Khanh Nguyen >> wrote: >>> I see.. >>> >>> What if I only need to resume the cluster and don't worry about any >>> data? Each time I use whirr to set up a new clusters, new EBS volumes >>> are also created and I got charged for the volumes at full month rate >>> even though I only use them for a few hours.Thanks. >>> >>> -k >>> >>> On Sun, Jun 5, 2011 at 5:43 PM, Andrei Savu wrote: >>>> Hi, >>>> >>>> There is no easy way to suspend a cluster started with Whirr without >>>> loosing all the data stored on the nodes. You need a custom tool to >>>> export and save data on shutdown (maybe to S3) and load it back when >>>> you start the cluster again. >>>> >>>> We are tracking progress on this issue here: >>>> https://issues.apache.org/jira/browse/WHIRR-309 >>>> >>>> Cheers, >>>> >>>> -- Andrei Savu / andreisavu.ro >>>> >>>> On Mon, Jun 6, 2011 at 12:34 AM, Khanh Nguyen >>>> wrote: >>>>> Hi, >>>>> >>>>> I followed the instructions here >>>>> http://nubetech.co/setting-up-cassandra-clusters-in-the-cloud-with-whirr >>>>> to set up an ec2 cluster on AWS. Everything works great until I stop >>>>> my instances (to save money) and restart them because the instances >>>>> get new IPs. What changes do I need to make in order to get my >>>>> cassandra cluster up and running again? Thanks. >>>>> >>>>> Regards, >>>>> >>>>> -k >>>>> >>>> >>> >> >
set up a cassandra cluster with ByteOrderedPartitioner using whirr?
Hi, I'm struggling to set up a cassandra cluster with ByteOrderedPartitioner using whirr. (I'm not sure if the issue is caused by Cassandra or Whirr so I cc-ed both lists). Here are the steps I took - use whirr to lauch a cassandra (version 0.8) cluster - ssh into each instances and do 1) kill cassandra 2) edit "partitioner" field in cassandra.yaml from "org.apache.cassandra.dht.RandomPartitioner" to "org.apache.cassandra.dht.ByteOrderedPartitioner" 3) edit 'JMX_PORT" in cassandra-env.sh from 7199 to 8080 (whirr bind JMX to port 8080) 3) delete "/var/lib/cassandra/data" 4) run cassandra again in the end, I got my cluster back but when I do 'describe cluster' inside cassandra-cli, it shows the cluster is still running RandomPartitioner. What am I missing? Thanks. Regards, -k
Re: ssh into instance started by whirr from ip address / machine different from the one that launched the cluster
Can you copy your .ssh/id_rsa and use it at the other machines? -k On Tue, Jun 7, 2011 at 6:52 PM, John Conwell wrote: > Is there anyway I can configure whirr to allow me to ssh into any node in my > cluster from a different machine than the one that launched the cluster? My > clusters can be fairly long lived, and I might need to access one of the > nodes via ssh from home, or customer's office, or someone else on my team, > or some such. Is there any way I can configure whirr to allow this?
set up two cassandra clusters?
Hi, I want to launch another cassandra clusters on EC2 but I keep getting an exception like this Exception in thread "main" java.lang.IllegalStateException: The permission '209.6.54.22/32-1-9160-9160' has already been authorized on the specified group Essentially, I am trying to launch two clusters. One uses RandomPartitioner and another use ByteOrderedPartitioner. Thanks. Cheers, -k
Re: set up two cassandra clusters?
Ah.. thank you. It works now. -k On Tue, Jun 7, 2011 at 8:11 PM, Tom White wrote: > Do the clusters have different names? Can you supply the stacktrace > you're getting from whirr.log. > > Cheers, > Tom > > On Tue, Jun 7, 2011 at 4:45 PM, Khanh Nguyen wrote: >> Hi, >> >> I want to launch another cassandra clusters on EC2 but I keep getting >> an exception like this >> >> Exception in thread "main" java.lang.IllegalStateException: The >> permission '209.6.54.22/32-1-9160-9160' has already been authorized on >> the specified group >> >> Essentially, I am trying to launch two clusters. One uses >> RandomPartitioner and another use ByteOrderedPartitioner. Thanks. >> >> Cheers, >> >> -k >> >
instances at different zones?
Hi, Is it possible to launch a cluster whose instances are at different zones? Thanks. -k
Re: instances at different zones?
I wan to compare the effect of communication load between nodes for different partitioning scheme . Having all instances in the same zone makes the latency between nodes too small to detect any significant difference. That's why I wanted to have instances in different zone. Anyway, I found a solution which uses netem to introduce a delay to all the connections out of the instances. Cheers. -k On Thu, Jun 9, 2011 at 1:36 AM, Andrei Savu wrote: > No. What do you want to accomplish? > > -- Andrei > > On Jun 9, 2011 7:12 AM, "Khanh Nguyen" wrote: >> Hi, >> >> Is it possible to launch a cluster whose instances are at different >> zones? Thanks. >> >> -k >
configure EBS volume size for ebs-instance
Hi, I used an ebs image (ami-1aad5273) to launch a Cassandra cluster and downloaded a 4Gb zip file from S3 to my instances. I ran into not enough disk error when I tried unzipping the data. How do I specify the size of the ebs volume in my configuration file? Thanks. Regards, -k
launching multiple services at once?
Hi, I know this feature not available yet but I was wondering if in the future, whirr can launch multiple services at once. Say I want to set up one cluster of 16 instances running both Hadoop and Cassandra. Thanks. -k
Re: launching multiple services at once?
Opps, thank you. I feel stupid :( -k On Sat, Jun 25, 2011 at 11:22 AM, Andrei Savu wrote: > You can already do that just by adding the roles as needed. > > -- Andrei Savu / andreisavu.ro > > On Sat, Jun 25, 2011 at 6:20 PM, Khanh Nguyen > wrote: >> Hi, >> >> I know this feature not available yet but I was wondering if in the >> future, whirr can launch multiple services at once. Say I want to set >> up one cluster of 16 instances running both Hadoop and Cassandra. >> Thanks. >> >> -k >> >