5 Essential Tips for Maximizing Your Experience at Nutanix .NEXT for Bloggers
Hi @AnishWalia20, now I understand better the speed difference between sftp and curl. For the moment is all. Thanks for the support
BTW: as explained somwhere on KB articles, now the images can be downloaded with sftp or curl.On my test i found that on 10Gb adapters the sftp is really slower than a curl api call: between 17 and 22 Mbit/s with sftp and close to 210 MBit/s with curl (same cluster, same vlan)# sftp testsftp -P 2222 user@CVM-IP#Connected to CVM-IP.sftp> cd /OsImages/.acropolis/vmdisksftp> get 43f7d8aa-712d-4bbe-a1e1-a10fa2501878Fetching /OsImages/.acropolis/vmdisk/f4c487e5-3597-431a-9e28-4bb5a17b14a1 to f4c487e5-3597-431a-9e28-4bb5a17b14a1/OsImages/.acropolis/vmdisk/f4c487e5-3597-431a-9e28-4bb5a17b14a1 100% 1986MB 17.5MB/s 01:53sftp> bye # curl testcurl --noproxy "*" --insecure -u pbackup:pbackup-password \ -GET --header "Accept: application/json" \"https://PE-IP-ADDRESS:9440/api/nutanix/v3/images/f4c487e5-3597-431a-9e28-4bb5a17b14a1/file" \-o /store/zzz-back-test-001-scsi2.raw \ % Total % Received % Xferd Average Speed Time Time Time Current
# get VM info: I got 2 scsi disks on storage container and one scsi disk on VG $ acli vm.get zzz-back-test-001zzz-back-test-001 { config { allow_live_migrate: True annotation: "test for vm backup using API" disk_list { addr { bus: "ide" index: 0 } cdrom: True device_uuid: "4af36e57-f7d8-4a60-8e99-ba9ab09ba06d" empty: True } disk_list { addr { bus: "scsi" index: 0 } container_id: 9 container_uuid: "c2960ea4-f12e-45cd-bc7a-32d6abdfc61a" device_uuid: "a480f6ca-dfe4-48c5-83d0-a3eb1ca0a20a" naa_id: "naa.6506b8d26ce248e9e41db36524bf4d07" source_vmdisk_uuid: "1efe564d-6304-4c1e-82ad-4c8186ce87e0" vmdisk_size: 10737418240 vmdisk_uuid: "b217d50d-4c13-47f0-a38c-e19c80cf7451" } disk_list { addr { bus: "scsi" index: 1 } container_id: 1008784 container_uuid: "39ba803c-8dcf-47d2-8ddb-a25a14fdd9bf" device_uuid: "85241664-0537-4f6e-8863-0cddd4e0
Wops…. my fault: is used the wrong storage container on my test….To apologize i´m going to post the command sequence I used to create my disk images both from storage containers and volume groups.
HI @AnishWalia20 you got reason. I already read that KB article, but let me try to repeat the whole sequence. In short I will post the used commands and the error.
Kindly whitelist the CVM and Virtual IP of the PE on Prism Central. Hi Rahul,you got the point. I have one internal proxy and I forget to add items to whitelist when the PC was reinstalled! Now works fine.Thanks a lot.-Piero
Hi all, I found the problem! Inside the Prims Element proxy configuration both http and https check box where flagged and all the http communications redirected to the proxy... I tested two solutions: - Removal of http check box (I don´t think that any http communication is required inside the cluster) -Add the http storage to the proxy whitelist. Both solution are working and now I can create images from PE/acli/REST API using http as source at the speed of light. Have Fun -Piero
Hello @HITESH0801 , ...Oh yeah! It´s work! Using the sftp command instead of scp I can get the images from command line: sftp -P 2222 user@CVM-IP:/container-name/.acropolis/vmdisk/image-or-disk-uuid Thanks for your answer, you saved my W.E. :D Kind regards -Piero
Hi David Good job! The availability of the source code would be great (the last I found on github was 1.5) Will you publish the code?
Hi [user=92275]Brian93[/user] I have contacted them but the technical support told me that VMs with attached Volume Groups are not supported.
Hi all, we are try to solve a problem related to VMs attached to Volume Groups (VG): Our Nutanix infrastructure at the moment is composed of 2 clusters based on AOS/AHY and cross-replicated using the native protection domains as DR. Excluding other standard VM services, we can say that the first cluster contain the OpenStack infrastructure (Database(s), Keystone, Horizon, etc.) and with the Nutanix OpenStack drivers we use the second cluster to generate the OpenStack VMs. Right now we haven´t still found a software solution that have VG support and we discovered this problem using Cohesity. For the moment negative answer support of vProtect and Veeam and we are waiting for other answers. Any suggestion?
Yep. You got reason... a wrong visualization on my VM :/
Thanks [user=39981]MMSW_DE[/user], yes I was talking about HYCU. But mow I'm a little bit confused because the [url=https://www.hycu.com/blog/how-hycu-uses-nutanix-volume-group-apis-for-consistent-data-protection/]HYCU BLog [/url]article say that Volume Groups are supported, but on the compatibility Matrix ([url=https://www.hycu.com/wp-content/uploads/2017/03/HYCU_CompatibilityMatrix.pdf]HYCU compatibility Matrix[/url] the iscsi is marked as N/A (as target) and the know that VMs generated from OpenStack internally use VG as iscsi devices. Anyway today I sent the question to both HYCU and Veeam. I will update the blog as soon I will find a solution.
Does this solution Volume Groups? We are using also OpenStack on Nutanix and is using VG[url=https://next.nutanix.com/sso/login?ssoType=saml] [/url] Tested Cohesity but doesn´t support VG and also have support for limited number of AOS versions :(
Already have an account? Login
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
Sorry, we're still checking this file's contents to make sure it's safe to download. Please try again in a few minutes.
Sorry, our virus scanner detected that this file isn't safe to download.
