Overview
This article addresses the issue of the SSI process being down in HOBO.
Solution
When the SPF cluster is completely down, it might cause the SSI process to only show in starting state.
Run the following procedure to reestablish the SPF NDB Cluster:
- Log in to both SPF's as ROOT on 2 separate windows:
- Run "systemctl stop mysql" on both.
- Run "systemctl stop ndbmtd" on both.
- Run "systemctl start ndbmtd" on both.
- Wait until NDB is loaded on both nodes. Check with the command "ndb_mgm -e show" and check for the 2 first lines (id=2 and id=3). They should be in a CONNECTED state (allow a few minutes for this to be reflected).
- Run "systemctl start mysql" on both nodes.
- Check again with "ndb_mgm -e show" and check for ids - 4, 5, 6, 7, 9, and 11.
- If everything up to this point is as expected, check the SSI state in both traffic servers.
Note: The procedure above will reload the NDB cluster only. You can run it without tp_stop / tp_start. If some errors still persist at the end of the procedure, then you can run tp_stop / tp_start to clear them (on SPF nodes only). No tp_stop / tp_start at the traffic nodes needed.