MemSQL tools settings to look for different port


#1

Hi,

I have a testing box where both the mysql and memsql are installed, since mysql default port is same as memsql we have installed memsql with different port. Now that I have upgraded my memsql to 6.7.18 and also installed the memsql management tools on it.

The issue I am facing is that, when using the memsql tools,I am not able to see the cluster(master/leaf nodes) however with memsql-ops commands I do see them

MemSQL Tools example:

    [17.1.0:vagrant@feature-114:1 ~]$ memsql-admin show-leaves
    failed to find host with master: no master aggregator found in this cluster

memsql-ops example:

[17.1.0:vagrant@feature-114:1 ~]$ memsql-ops memsql-list
 ID       Agent Id  Process State  Cluster State  Role    Host       Port   Version 
 66242E5  A37f783   RUNNING        CONNECTED      MASTER  127.0.0.1  43306  6.7.18  
 3CCD328  A37f783   RUNNING        CONNECTED      LEAF    127.0.0.1  43307  6.7.18

#2

To start using MemSQL Tools with an existing cluster, you need to run https://docs.memsql.com/memsql-tools-reference/latest/register-node/ to register each node with Tools.


#3

Even after registering I am still getting the same issue

[17.1.0:vagrant@feature-114:0 lib]$ memsql-ops memsql-list
ID Agent Id Process State Cluster State Role Host Port Version
66342E5 A37f783 RUNNING CONNECTED MASTER 127.0.0.1 43306 6.7.18
3C2D328 A37f783 RUNNING CONNECTED LEAF 127.0.0.1 43307 6.7.18
[17.1.0:vagrant@feature-114:0 lib]$ memsql-toolbox-config list-hosts
±----------±-----------±------------±--------------+
| Host | Local Host | SSH address | Identity File |
±----------±-----------±------------±--------------+
| 127.0.0.1 | Yes | | |
±----------±-----------±------------±--------------+
[17.1.0:vagrant@feature-114:0 lib]$ memsql-admin list-nodes
✘ Failed to list nodes on all hosts: failed to list nodes on 1 host: 127.0.0.1
No nodes found
[17.1.0:vagrant@feature-114:0 lib]$

I have tried the other way aswell

[17.1.0:vagrant@feature-114:0 lib]$ memsql-admin register-node --host 127.0.0.1 --port 43306 --memsql-config /var/lib/memsql/master-43306-MIa4c4bbc0/memsql.cnf

could not determine memsqlctl user: error running memsqlctl: failed to start command: "memsqlctl" "--json" "--yes" "env": exec: “memsqlctl”: executable file not found in $PATH

stderr:


Memsql-toolbox-config register-host doesn't look for loopback IP or port
#4

@jack Thanks , can you please suggest solutions what I have to follow for the upgrade.

List of issues:

We have AWS image where we create and tear it off very often, this AWS image is just a single host and memsql is installed with different ports(43306 master,43307 leaf). I have successfully upgraded with memsql ops.

We have later installed memsql tools with which all the issues I am facing with.

  1. We would like to register this IP(127.0.0.1) common for every AWS EC2 created irrespective of its original IP . This was easier through memsql-ops but not with memsql tool command line we are facing issues using common IP for master and leaf.
  2. Even with registering I couldn’t add nodes (at one point I have completely uninstalled and followed the document but then I was able to register and point the master node but failed with leaf node)
  3. I think the problem seems to be with memsqlctl and nodes.hcl, is there something we need to add these files manually and add some config in it.