VM Creation #11852
closed
Create a VM to deploy the new storage hub application.
100%
all
storagehub latest version
storagehub
120GB under /data
40 GB under /tomcat/temp
Description
I need to install the storagehub webapp on this VM but it seems to be not reachable:
rcirillo@rcirillo-cnr:~$ ping workspace-repository-prod.d4science.org PING workspace-repository-prod.d4science.org (146.48.122.106) 56(84) bytes of data. ^C --- workspace-repository-prod.d4science.org ping statistics --- 11 packets transmitted, 0 received, 100% packet loss, time 10079ms
Could anyone take a look, please?
Related issues
Updated by Roberto Cirillo almost 7 years ago
- Related to Upgrade #11834: D4Science Root VO upgrade to gCube 4.11.1 added
Updated by Andrea Dell'Amico almost 7 years ago
That VM was destroyed because it was unused. If there are plans to finally activate the cluster, the VM must be reinstalled.
Updated by Roberto Cirillo almost 7 years ago
Andrea Dell'Amico wrote:
That VM was destroyed because it was unused. If there are plans to finally activate the cluster, the VM must be reinstalled.
There was a ticket for this activity? I haven't see it. Is it possible to create another VM with the same resources in order to install the new storagehub webapp?
Updated by Andrea Dell'Amico almost 7 years ago
- Related to Task #10264: Add /data as separate file system on workspace-repository-prod.d4science.org added
Updated by Andrea Dell'Amico almost 7 years ago
Not a dedicated one, but the matter was discussed in #10264 that I just added as related.
Please open a new ticket or change this one in VM creation
Updated by Andrea Dell'Amico almost 7 years ago
Also keep in mind that we will need a new load balancer configuration.
Will the new storage hub able to work in a cluster?
Updated by Roberto Cirillo almost 7 years ago
- Tracker changed from Support to VM Creation
- RAM set to 8
- CPU set to 16
- DiskSpace set to 30
- Smartgears set to Yes
- Smartgears Version set to latest
- Smartgears scopes updated (diff)
- Applications list and version updated (diff)
- Applications context path updated (diff)
- External storage data disk updated (diff)
- Number of VMs set to 1
This new VM shouldn't be in the same cluster with the workspace-repository-prod1.d4science.org VM but it should share the same db (workspace-repository-pg) and JR configuration.
Please @lucio.lelii@isti.cnr.it check if the resources in the description are ok for you. I've just set the same resources of the old workspace-repository-prod.d4science.org VM.
The new VM name could be the following if you are agree:
storagehub.d4science.org
Updated by Andrea Dell'Amico almost 7 years ago
I think it's better if we talk about it.
Updated by Massimiliano Assante almost 7 years ago
I don't think we should rename the VM, let's keep it workspace-repository-prod.d4science.org
@lucio.lelii@isti.cnr.it are we sure that This new VM shouldn't be in the same cluster with the workspace-repository-prod1.d4science.org VM but it should share the same db (workspace-repository-pg) and JCR configuration.
Updated by Lucio Lelii almost 7 years ago
Massimiliano Assante wrote:
@lucio.lelii@isti.cnr.it are we sure that This new VM shouldn't be in the same cluster with the workspace-repository-prod1.d4science.org VM but it should share the same db (workspace-repository-pg) and JCR configuration.
Yes
Updated by Massimiliano Assante almost 7 years ago
Lucio Lelii wrote:
Massimiliano Assante wrote:
@lucio.lelii@isti.cnr.it are we sure that This new VM shouldn't be in the same cluster with the workspace-repository-prod1.d4science.org VM but it should share the same db (workspace-repository-pg) and JCR configuration.
Yes
then the name proposed by Roberto makes sense.
Updated by Roberto Cirillo almost 7 years ago
I would also like to say that for the moment, the storagehub is in ready-only mode on JR repository.
Updated by Andrea Dell'Amico almost 7 years ago
I would keep the workspace-repository.d4science.org hostname on the balancer to overcome the certificates problem. The app context is different, so the load balancer can discriminate on that.
Updated by Roberto Cirillo almost 7 years ago
- Related to deleted (Upgrade #11834: D4Science Root VO upgrade to gCube 4.11.1)
Updated by Roberto Cirillo almost 7 years ago
- Blocks Upgrade #11834: D4Science Root VO upgrade to gCube 4.11.1 added
Updated by Massimiliano Assante almost 7 years ago
- Due date set to May 31, 2018
it would be beneficial if the VM would be up by end of today? The sooner the better
Updated by Massimiliano Assante almost 7 years ago
@andrea.dellamico@isti.cnr.it @tommaso.piccioli@isti.cnr.it could you update the status of this activity please? Thanks!
Updated by Andrea Dell'Amico almost 7 years ago
- Status changed from New to In Progress
I'm going to name it storagehub1.d4science.org.
Updated by Andrea Dell'Amico almost 7 years ago
- % Done changed from 0 to 80
The VM is up. I've added the two volumes, they will be mounted by the playbook.
I also cleaned up the playbook and added a specific playbook for the storage hub: storage-hub.yml (in the long run the smartgears-nodes.yml playbook should disappear, it's too bloated).
The storage hub variables now have a common group named storage_hub, and then specific groups for the prod and dev environments. I did not cleaned up the dev variables.
@roberto.cirillo@isti.cnr.it you have root access, you can run the playbook.
I still have to change the load balancer configuration.
Updated by Roberto Cirillo almost 7 years ago
Thank you @andrea.dellamico@isti.cnr.it I'm going to run the playbook right now
Updated by Roberto Cirillo almost 7 years ago
The following task is failing:
TASK [../../library/roles/data_disk : Create a file system on the new disks] **********************************************************************************************************************************************************
fatal: [storagehub1.d4science.org]: FAILED! => {"msg": "'catalina_tmp_directory' is undefined"}
Please, could you check?
Updated by Andrea Dell'Amico almost 7 years ago
Wrong variable, now it should be fixed.
Updated by Andrea Dell'Amico almost 7 years ago
The load balancer has been configured. The main URL is workspace-repository.d4science.org ( /storagehub,/homelibrary-fs-webapp).
Updated by Andrea Dell'Amico almost 7 years ago
- Subject changed from workspace-repository-prod.d4science.org: is unreachable to Create a VM to deploy the new storage hub application.
- % Done changed from 80 to 90
The VM has been provisioned. A working check for the load balancer is needed, because /storagehub returns an error even if the application has been deployed. @lucio.lelii@isti.cnr.it can you give me a working URL?
Updated by Lucio Lelii almost 7 years ago
the check to understand if the node is up is {hostname}/storagehub/gcube/resource/
Updated by Andrea Dell'Amico almost 7 years ago
Done. Now the load balancer is happy.
Updated by Lucio Lelii almost 7 years ago
It isn't working, I tried the following URL
http://workspace-repository.d4science.org/storagehub/gcube/resource/
but it returns a 404 error
Updated by Andrea Dell'Amico almost 7 years ago
I forgot to also change the ACL used by the old home library. Your request were intercepted by that rule and redirected to the wrong server. Now it works.
Updated by Roberto Cirillo almost 7 years ago
I've added "workspace-repository.d4science.org" as smartgears_hostname. Now the storagehub service is published as the haproxy address
Updated by Roberto Cirillo almost 7 years ago
- Status changed from In Progress to Closed
