10.5 Monitoring the resource usage (list of messages to be monitored)
This section describes the messages that need to be monitored when you monitor the usage of resources used by the HADB server.
The HADB server runs by using various resources. Therefore, if any of the following events occur, the operational status of the HADB server might be affected:
-
The resources of the server machine or the OS (such as disks, CPU, or memory) are insufficient.
-
The resources managed by the HADB server (such as the numbers of connections and chunks) are insufficient.
-
Failure occurs in a disk or the network.
-
The lock control causes a wait for processing.
If such an event occurs, you need to promptly take the appropriate actions. To do this, the HADB administrator needs to check the message output by the HADB server, and monitor the resource usage.
The following table lists and describes the resources used by the HADB server and the messages to be monitored. For details about how to handle individual output messages, see the manual HADB Messages.
No. |
Resource type |
Message to be monitored |
Description |
|
---|---|---|---|---|
1 |
Locked resource |
|
If one of the messages in the left column is output, the process of allocating a locked resource is placed in wait status. For details about how to handle the message, see the column that indicates the handling of the applicable message. For details about how to check whether the process of allocating a locked resource has gone into wait status, see 10.8.3 Checking whether the process of reserving locked resources has gone into wait status. If the process of allocating a locked resource repeatedly goes into wait status, the application or command that allocated the locked resource might have been running for a long time. Also, see 10.8.1 Checking the application or command processing status. |
|
2 |
Memory |
Memory managed by the HADB server |
|
The KFAA40021-W message is a warning message that is output when the memory usage on the HADB server is no less than the ratio specified for the adb_sys_memory_limit_wrn_pnt operand in the server definition. (This operand indicates the trigger for outputting the warning message regarding the memory usage on the HADB server.) If the KFAA40021-W message is output, check the memory usage on the HADB server. For details about how to check the memory usage, see 10.6.3 Checking the memory usage status for each real thread. The KFAA40024-I message is an information message that is output when the memory usage on the HADB server is no more than the ratio specified for the adb_sys_memory_limit_wrn_pnt operand in the server definition. (This operand indicates the trigger for resetting the status indicating that a warning message has been output.)
|
3 |
|
If one of the messages in the left column is output, the memory managed by the HADB server is insufficient. When you execute an SQL statement or command, if memory shortage occurs, any of the following messages is output: When SQL statement is executed
When a command is executed
For details about how to handle the above output messages, see the column that indicates the handling of KFAA40007-E message. In addition, regularly execute the adbstat command to check the usage status of the memory used by the HADB server. (Check whether memory usage is near the upper limit.) For details, see 10.6.1 Checking the usage status of all memory. |
||
4 |
Memory used by the libraries that the HADB server uses |
|
If one of the messages in the left column is output, the memory used by the libraries that the HADB server uses is insufficient. When you execute a command, if memory shortage occurs, any of the messages in the left column is output. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
5 |
Shared memory |
KFAA40002-E |
If the message in the left column is output, an error occurred when a shared memory was accessed. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
6 |
CPU |
None |
The HADB server does not output messages about CPU usage rate. To monitor the CPU usage rate of the HADB server, execute the following OS commands:
|
|
7 |
Inter-process communication |
|
If one of the messages in the left column is output, an error occurred in inter-process communication of the HADB server. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
8 |
Directory |
DB directory |
KFAA30959-E |
If the message in the left column is output, an error occurred when the DB directory was accessed. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
9 |
Archive directory |
KFAA31666-E |
If the message in the left column is output, an error occurred when the archive directory was accessed. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
10 |
KFAA61212-W |
If the message in the left column is output, an error occurred in the processing for deleting the archive directory. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
||
11 |
File |
Message log file |
KFAA40025-W |
If the message in the left column is output, one of the following message log files is in the fall-back mode:
If the message in the left column is output, release the fall-back mode of the message log file. For details about how to release the fall-back mode, see 15.16.1 Releasing the message log file from fall-back mode. The message in the left column is output to syslog even if the message cannot be output to the message log file. Therefore, you need to monitor both the message log file and syslog. |
12 |
KFAA50057-W |
Multi-node function If the message in the left column is output, the message log file on a node is in the fall-back mode. In this case, release the fall-back mode of the message log file. For details about how to release the fall-back mode, see 15.16.1 Releasing the message log file from fall-back mode. |
||
13 |
DB area file |
|
If the message in the left column is output, operation on the applicable file failed. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
14 |
Archive file |
|||
15 |
System file |
|||
16 |
CSV file whose information is input to the ADB_CSVREAD function |
|||
17 |
File on which the command is executed |
|
||
18 |
HADB dump file |
|
||
19 |
Message catalog file |
|
||
20 |
Shared memory ID storage file |
KFAA50028-E |
||
21 |
Shared memory dump file |
KFAA50029-E |
||
22 |
Semaphore set ID storage file |
KFAA50037-E |
||
23 |
Command status file |
KFAA50243-E |
||
24 |
Temporary work file |
KFAA50245-E |
||
25 |
Index record file |
KFAA50246-E |
||
26 |
SQL trace file |
KFAA51010-W |
||
27 |
Synonym dictionary file |
|
||
28 |
KFAA51537-W |
Multi-node function If the message in the left column is output, the latest synonym dictionary file has not been stored in the directory specified for the adb_syndict_node_storage_path operand in the server definition. In this case, performance of synonym search might be deteriorated. For details about how to handle the message, see the column that indicates the handling of that message. |
||
29 |
Unload file |
KFAA61400-W |
If the message in the left column is output, deletion of the unload file that was temporarily created when the adbreorgsystemdata command was executed failed. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
30 |
Audit trail file |
KFAA51404-E |
If the message in the left column is output, an operation on an audit trail file has failed. For details about how to handle the message, see the column that indicates the handling of that message. |
|
31 |
KFAA51411-W |
The message in the left column is output if the current audit trail file cannot be found when attempting to swap the audit trail file. The audit trail information output to the current audit trail file might have been lost. |
||
32 |
KFAA81401-I |
The message in the left column is output when the audit trail file has been swapped. If you perform any of the following operations for the audit trail file, monitor the message in the left column:
|
||
33 |
KFAA81402-I |
The message in the left column is output when the audit trail file has been renamed. If you perform any of the following operations for the audit trail file, monitor the message in the left column:
|
||
34 |
File used by the updated-row columnizing facility |
KFAA41220-W |
If the message in the left column is output, there is an error in a file that is used for the updated-row columnizing facility. For details about how to handle the message, see the column that indicates the handling of that message. |
|
35 |
Node |
|
If one of the messages in the left column is output, an abnormality occurs in the node on which the message is output. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
36 |
DB area |
|
If one of the messages in the left column is output, the free disk space is insufficient, and data cannot be stored in the target DB area. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
37 |
Log file |
|
If one of the messages in the left column is output, the number of user log files is insufficient. For details about how to handle the message, see the column that indicates the handling of the applicable message. Note that, if you omit specifying the adb_log_usrfile_num operand in the server definition, an error due to insufficient user log files will not occur. For details about the adb_log_usrfile_num operand in the server definition, see the description of the adb_log_usrfile_num operand in 7.2.3 Operands related to system logs (set format). |
|
38 |
KFAA61211-E |
If the message in the left column is output, the free space of the disk in which the system log file is stored is insufficient. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
||
39 |
KFAA81215-I |
If the message in the left column is output, database recovery processing is being performed. For details about how to handle the message, see the column that indicates the handling of that message. The message in the left column is output at intervals specified for the adb_log_rec_msg_interval operand in the server definition until the database recovery processing is complete. For details about the adb_log_rec_msg_interval operand in the server definition, see the description of the adb_log_rec_msg_interval operand in 7.2.3 Operands related to system logs (set format). |
||
40 |
Buffer |
|
If one of the messages in the left column is output, the number of buffers is insufficient. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
41 |
|
If one of the messages in the left column is output, the load on the buffer is too high. For details about how to handle the message, see the column that indicates the handling of the KFAA41202-E message. |
||
42 |
Chunk |
KFAA51245-E |
If the message in the left column is output, a new chunk cannot be created because the number of chunks will exceed the maximum number. The number of chunks that can be created in a DB area is about to exceed the maximum number. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
43 |
KFAA51246-E |
If the message in the left column is output, any new chunk cannot be created because the number of chunks will exceed the maximum number. The number of chunks that can be created in a table is about to exceed the maximum number. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
||
44 |
Hash table area |
KFAA51130-W |
If the message in the left column is output, a work table has been created because the hash table area was insufficient. For details about how to handle the message, see the column that indicates the handling of the applicable message. Also, check whether any problems occur when the hash table area is insufficient. (Check whether the situation in which the hash table area is insufficient is assumed in the database design.) |
|
45 |
Connection |
|
The KFAA40020-W message is a warning message that is output when the number of connections to the HADB server is no less than the ratio specified for the adb_sys_max_users_wrn_pnt operand in the server definition. (This operand indicates the trigger for outputting the warning message regarding the maximum number of concurrent connections.) If the KFAA40020-W message is output, check whether there is an application or command that is illegally connecting to the HADB server. For details about how to check this, see 10.8.1 Checking the application or command processing status. The KFAA40023-I message is an information message that is output when the number of connections to the HADB server is no more than the ratio specified for the adb_sys_max_users_wrn_pnt operand in the server definition. (This operand indicates the trigger for resetting the status indicating that a warning message has been output.) If the client-group facility is being used, instead of the value specified for the adb_sys_max_users_wrn_pnt operand, the following values specified for the -w option of the adbcltgrp operand are applied:
|
|
46 |
KFAA30932-E |
If the message in the left column is output, an application or command cannot connect to the HADB server. The number of connections to the HADB server exceeds the maximum number of concurrent connections. For details about how to handle the message, see the column that indicates the handling of the applicable message. Check whether the value specified for the adb_sys_max_users operand in the server definition is appropriate. For details about the adb_sys_max_users operand in the server definition, see the description of the adb_sys_max_users operand in 7.2.1 Operands related to system configuration (set format). |
||
47 |
Statement handle |
KFAA30931-E |
If the message in the left column is output, the number of statement handles exceeds the upper limit. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
48 |
Base table |
|
If one the messages in the left column is output, the number of base tables or indexes that can be defined in the system exceeds the upper limit. Otherwise, the number of base tables or indexes that can be stored in the target DB area exceeds the upper limit. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
49 |
Index |
|||
50 |
Viewed table |
KFAA30811-E |
If the message in the left column is output, the number of viewed tables that can be defined in the system exceeds the upper limit. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
51 |
system table (base table) |
|
If one of the messages in the left column is output, there are too few unused segments in the system-table DB area. Therefore, you need to execute the adbreorgsystemdata command to reorganize the system table. For details about how to handle the message, see the column that indicates the handling of the applicable message. |
|
52 |
HADB user |
KFAA30811-E |
If the message in the left column is output, the number of HADB users that can be defined in the system exceeds the upper limit. For details about how to handle the message, see the column that indicates the handling of the applicable message. |