14.3.25 resume (resumes caches)
This subsection is applicable when you are using disk caches or two-way caches.
- Organization of this subsection
(1) Description
This subcommand resumes disk caches and two-way caches.
When this subcommand is executed, the caches that have both cache property files and cache information files are resumed.
Use this subcommand to resume caches in the following cases:
-
For restarting the EADS servers after the cluster was stopped normally
-
For restoring the cluster after it has been in unavailable (NOT_AVAILABLE) or partially available (PARTIALLY_AVAILABLE) status
(2) Rules
-
This subcommand can be executed only when the status of the cluster is the following:
-
Cluster available (AVAILABLE)
-
-
The target of this subcommand is the EADS servers whose cluster participation status is online. This subcommand cannot be executed if the cluster contains any EADS server whose cluster participation status is offline or standby. You can determine the cluster participation status with the eztool status command.
-
This subcommand can be executed when the target EADS servers are in the following status:
-
Initialized
-
-
To prevent a full garbage collection (FullGC) from occurring while operations are underway, each EADS server performs FullGC when this subcommand terminates.
-
The EADS servers are not isolated while this subcommand has obtained a lock from the EADS servers. However, if the process is shut down or an EADS server terminates while the subcommand has the lock, the lock is released, and then the EADS server becomes isolated.
(3) Format
eztool resume
(4) Return code
The following table lists the return codes that this subcommand returns.
No. |
Return code |
Description |
|
---|---|---|---|
eads.command.compat parameter omitted in the command properties |
0300 specified in the eads.command.compat parameter in the command properties |
||
1 |
0 |
0 |
Command execution was successful. |
2 |
101 |
101 |
Initialization of the command failed. |
3 |
110 |
Connection establishment failed. |
|
4 |
111 |
The command failed due to a communication timeout. |
|
5 |
120 |
The command failed due to a syntax error. |
|
6 |
130 |
The command failed because it could not be executed. |
|
7 |
131 |
The command failed because another command was executing. |
|
8 |
150 |
The command failed during execution. |
|
9 |
200 |
The command failed due to a timeout. |
(5) Notes
-
Make sure that the properties of the caches that will not be resumed are also identical on all EADS servers that make up the cluster (except for the eads.cache.disk.info.dir and eads.cache.disk.n.dir parameters).
-
If the cluster was not stopped normally, deleted data might be restored.
-
If the cache resume processing terminates prematurely for a reason such as abnormal termination or a command timeout, locks are not released. If locks have not been released after an attempt has been made to resume caches, data might no longer be consistent. If you continue operation in such a status, data might become corrupted or lost. To avoid this, take the following steps:
-
Execute the eztool status -v command to check the cluster status.
-
Execute the eztool unlock command to release the lock.
-
Execute the eztool listcache command to check the list of caches.
-
If the cache resume processing has failed, execute the eztool deletecache command to delete the caches.
-
If EADS servers are isolated, restore the EADS servers.
-
Perform cache resume processing again.
-
-
If the cache resume processing has failed, check the following items and then perform the cache restart processing again:
-
Check for errors in the parameters specified in the cluster property files or shared property files.
-
Check for errors in the parameters specified in the cache property files.
-
Check the cache files for any invalid status.
-
-
The total data restriction function is disabled while this subcommand is executing.