CODEWITHSUNDEEP
amazon-ec2datastax-enterpriseamazon-amidatastax
Alexey Ossikine
Alexey Ossikine

Reputation: 113

DataStax Enterprise AMI hangs during installation

I booted the DataStax AMI for Amazon EC2, logged in via SSH, but the terminal hangs on "Installation Started":

Cluster started with these options:
--clustername CassandraDev --totalnodes 1 --version enterprise --username **** --password ****

Installation started.

"Installation started" keeps going through suffixes consisting of one, two, and three dots. But nothing happens, I can't quit the installation process, and I can't access any log files to see what might be going on (or I don't know how).

Tried on two separate m3.large instances operating in a VPC subnet, at the us-east-1 region. The exact AMI is datastax_clustering_ami_2.5.1_hvm.manifest.xml (ami-ada2b6c4). On the first instance, I waited about an hour and a half. The second instance I just left online all night, with the same results.

Because this is a VPC, all outbound traffic goes through a NAT server. Security groups allow outbound traffic only on ports 80, 443, and 123. Might there be another outbound port that needs to be opened? Inbound ports do not matter, as the server is not public-facing, but within the subnet I have allowed all traffic on all ports.

Someone else has had a similar issue, but without answers so far: DataStax AMI hangs on

Any help would be appreciated!

Upvotes: 0

Views: 306

Answers (1)

joaquincasares
joaquincasares

Reputation: 113

Since there were a few tickets that came up recently around the same issue, it seems as though something recently changed within the AMI provisioning side in EC2, or this specific configuration of VPCs had never been used before, which seems a bit unlikely.

The current fix is to add an additional entry into /etc/hostname to get rid of the stderr output that occurs after each sudo command. This in turn doesn't get flagged as an error on the provisioning side.

This has been fixed and patched as documented on this ticket: https://github.com/riptano/ComboAMI/issues/51.

If you spot any additional issues, feel free to create another ticket there.

Going forward, just launch another set of instances using the same user-data and you should be up and running.

Upvotes: 1

Related Questions