Causes		Symptoms	Countermeasures
	Category 1	Category 2
	Process cannot start normally	Recovery failure	[Error code] From 20000 [Description] Recovery-related error group [Message] In accordance with the format of the recovery error category	Possible causes of the error could be mistakes in the operating procedure or internal faults. See “Problems related to recovery process” for the details.
		Error in the contents described in the definition file (gs_cluster.json, gs_node.json)	[Server] 100002, 100003, 100004, 100005 [Description] Configuration setting failure [Message] Regarding the initial error which detected the failure, the failure cause will be displayed, be it the omission of an essential item, the value lying outside the limit range, or a discrepancy error in the data type.	Error in the gs_node.json/gs_cluster.json settings. Check the error message and manual to see if there is any omissions of essential items in the definition file, value limits and data type discrepancy issues and then correct them before restarting.
		Service address described in gs_node.json failed to resolve DNS name.	[Server] 130000 [Description] Service address setting failure [Message] Platform error no. resulting in failed address, port, error (in many cases -2 (EAI_NONAME))	Check whether each serviceAddress in gs_node.json is normal or not. If it has not been set, the address listed in /etc/hosts will be used so check this as well.
		Error in the multicast address described in gs_cluster.json	[Server] 130006 [Description] Multicast address setting failure [Message] Platform error no. resulting in failed address, port, error (22 (EINVAL))	Check whether each notificationAddress of gs_cluster.json is normal (224.0.0.0 to 239.255.255.255) or not. It is recommended that the port be split but this will work even if it is duplicated.
		Multicast cannot be executed due to a defect in the environmental settings.	[Server] 130006 [Description] Multicast address setting failure [Message] Platform error no. resulting in failed address, port, error (19 (ENODEV))	Initially, check whether a valid network connection other than a loopback (lo) exists, and whether it is valid or not. Next, check the multicast routing setting. If it is not specified in particular, the default gateway setting will be used, but check whether this setting is correct or not.
		Server and database versions do not match	[Server] 80022 [Description] Database version discrepancy [Message] Module Row no. Error code Log file version	Check the version no. stated in the error message, and transit to a database with a valid version. See “Compatibility between client and DB” for details regarding valid versions and the “GridStore migration procedure” for the migration method.
	During the process start-up, the node status changes to ABNORMAL.	One of the ports in gs_node.json is duplicated in this file or it is the same as the port of a process that is already being started by the same machine.	[Server] 130024 [Description] Socket bind failure [Message] Failed address, port, time required for retry (1 minute by default) [External tool] Use netstat, port scan, etc. to check for duplication and search for empty ports.	Use an external tool, etc. to check that the port listed in gs_node.json is not covered, and restart after correcting it to an appropriate value.
	Cluster is not composed even after a fixed period of time has passed.	Discrepancy in the values of gs_cluster.json or the cluster name and number of nodes constituting a cluster, among the nodes constituting the cluster.	[Server] 180046 to 180050 [Description] Cluster configuration failure [Message] Reason why connection destination address no., cluster could not be composed (cluster name discrepancy, configuration discrepancy, etc.)	Check that all values of gs_cluster.json as well as the cluster name and number of nodes constituting a cluster that are specified in gs_joincluster are correct among all the constituting nodes.
		Discrepancy in the valid binary version of the server among clusters	[Server] 40044 [Description] Server version discrepancy [Message] Connection destination address no., own server version no., connection destination server version no.	Check the version no. listed in the error, and use a valid version of the server binary to restart the cluster and recompose the cluster again.
		Master node of current cluster detected that another cluster with the same cluster name already exists (with the same multicast address).	[Server] 180018 [Description] Duplicated cluster detection [Message] Connection address, port no. of master node in an existing cluster	As data may be destroyed if the system continues to operate even after detecting a duplication, either stop the cluster on the side which detected the duplication (gs_stopcluster), or stop the cluster at the notification destination (repeat this if there are 2 or more clusters), and make adjustments so that clusters with the same cluster name become one on the same subnet. When setting the cluster name of gs_cluster.json, it is strongly recommended that a unique cluster name (and multicast address) be set.
		Service address in gs_node.json is not valid	[Server] Although no traces remain in particular, a cluster of 2 or more nodes cannot be composed even after a fixed period of time has passed	Check the legitimacy of the address of each service. Service will start if there is only 1 node but if there are 2 or more nodes, the cluster cannot be composed as communications cannot be carried out among the nodes.
		Multicast distribution to compose a cluster does not physically arrive.	[External tool] Check tcpdump and netstat to see whether the multicast address and port specified in the notification port and /cluster/notificationAddress of gs_cluster.json are being received.	Check that all the nodes constituting a cluster have been started in the same subnet. If a connection between different subnets is necessary, use an external tool to connect. In addition, check the settings on the other end as well since the relevant multicast may have been cut off by a firewall.
	Failure to connect client to cluster	Crossed wires at the connection destination due to clusters having the same cluster name in the same network	[Server] Cluster is uncertain due to failure to get container, etc., but in this case, the following is recorded by either one of the nodes in the cluster [Server] 10005 [Description] Duplicated cluster detection [Message] Connection address, port no. of master node in an existing cluster	First, check that the cluster name in unique in the same subnet. Then, check that the cluster name is the same as the name of the cluster to be connected to the client program.
		Binary version is not inconsistent between the client and cluster	[Server] 10054 [Description] Client version discrepancy [Message] Connection destination address no., version no. of own client that can be connected, connection destination client version no.	Check the version no. stated in the error message, and use a client with a valid version. See “Compatibility between client and DB” regarding the valid versions.
		A client failover timeout has occurred.	[Server] 10008, 10009, 10010 [Description] Failover connection failure [Client] Client failover timeout	This trace may be recorded each time a retry is attempted by a client. The log above will be left behind if a connection had been established during the client failover timed out, but this is not a problem. If a timeout has occurred, the last failure cause will be noted down. See “Problems related to client failover”.
		Multicast distribution for cluster discovery of a client does not physically arrive.	[External tool] Check tcpdump and netstat to see whether the multicast address and port specified in the notification port and /transaction/notificationAddress of gs_cluster.json are being received.	Check that the client have been started in the same subnet as the subnet in which the cluster exists. If a connection between different subnets is necessary, use an external tool to connect. In addition, check the settings on the other end as well since the relevant multicast may have been cut off by a firewall.
	Database version discrepancy	Server and database versions do not match	[Server] 80022 [Description] Database version discrepancy [Message] Module Row no. Error code Log file version	Check the version no. stated in the error message, and transit to a database with a valid version. See “Compatibility between client and DB” for details regarding valid versions and the “GridStore migration procedure” for the migration method.
	See “Annex”, “Parameter List” in the “GridStore quick start guide” for the gs_node.json and gs_cluster.json settings.
	A “Migration procedure” is provided in basic support services. Note that the procedure is not enclosed with the installation media and package.