Adding a MemSQL file
Unpacking archive to determine its version
Extracting utils/build_metadata.json
Fall back to utils/build_receipt.json
Error while extracting version information
Please check that the file at path /root/memsql-server-7.0.16-5aa7127018.x86_64.tar.gz is a MemSQL file
memsql-deploy upgrade --version 7.0
refusing to upgrade a cluster with no master aggregator, make sure that a master aggregator exists in this cluster and it shows up in ‘memsql-admin list-nodes’ before continuing
memsql-deploy upgrade --file-path /root/memsql-server.rpm
refusing to upgrade a cluster with no master aggregator, make sure that a master aggregator exists in this cluster and it shows up in ‘memsql-admin list-nodes’ before continuing
Thanks for sharing info on the error you are getting. For this cluster, was it deployed using Ops or Tools?
The tools error message is saying there is not a master aggregator present, which could be that your nodes are not registered with tools if they were originally deployed by Ops.
We do support offline upgrade from 6.x to 7.0 with Ops, so if the cluster was originally deployed with Ops we can proceed with troubleshooting that path, just let me know how you originally deployed the cluster.
Please send me the steps in detail ,so that i can try and implement them and get back to you in case of any concerns, but not even a single step is working as per document , i tried them the all possible ways
Forgot to tell you one more issue, when trying to do the same with ops even from 6.7 to 6.8 as the ops is upgraded to 7 its throwing error like can’t upgrade to 6.8 as the ops is in version 7.0.16
I apologize for the confusion. We have docs bug there.
The error your getting is because of the “–version 7.0”. You’ll need to specify the exact version you want if you use --version. If you skip --version, ops will download the latest version. So the command you want is:
memsql-ops memsql-upgrade
or memsql-ops memsql-upgrade --version 7.0.16
Thank you very much for the info, yes it worked, but when we are trying to install memsql 6.7 version using install.sh what its doing is its installing the latest version of MEMSQL V7
Only if we do manual process of using add file -t 6.7.14.tar.gz its installing 6.7 version where in we need to add the master and the leaf separately in those steps
Can you ask the concern team to update the UPGRADE doc
Hi Adam,
When i am trying to upgrade MEMSQL from 6.7.14 to V7 using Tools its not allowing me do
Its throwing the following error
./memsql-deploy upgrade
refusing to upgrade a cluster with no master aggregator, make sure that a master aggregator exists in this cluster and it shows up in ‘memsql-admin list-nodes’ before continuing
After this i tried migrating ops to tools and try upgrade then its throwing me the following error
memsql-deploy upgrade --version 7.0.16
The selected version is already installed on all hosts
-bash-4.2$ memsql-admin list-nodes ±-----------±-------±------------±-----±--------------±-------------±--------±---------------±-------------------±-------------+
| MemSQL ID | Role | Host | Port | Process State | Connectable? | Version | Recovery State | Availability Group | Bind Address |
±-----------±-------±------------±-----±--------------±-------------±--------±---------------±-------------------±-------------+
| 3D65234619 | Master | 10.128.0.34 | 3306 | Running | True | 6.7.14 | Online | | 0.0.0.0 |
| 5018587A0C | Leaf | 10.128.0.34 | 3307 | Running | True | 6.7.14 | Online | 1 | 0.0.0.0 |
| 6654639DB1 | Leaf | 10.128.0.34 | 3308 | Running | True | 6.7.14 | Online | 1 | 0.0.0.0 |
±-----------±-------±------------±-----±--------------±-------------±--------±---------------±-------------------±-------------+
Thanks for the work around , it worked but as i am doing it in test environment with only one node however do we need to perform the same in PROD Cluster as well with more nodes in place ?
Do we need to perform the un install and install on all nodes locally ?
Can you also suggest me the --file-path command as well
Do we have any chance of getting the detailed steps of doing upgrade on a multi node cluster
or
Can we expect an inclusion of detailed steps in the upgrade guide documentation ?