Patent Grant
7631002
The present invention relates to management technique of storage devices in which a use situations of the storage devices utilizing a storage area network (SAN) are predicted to manage the storage devices on the basis of the predicted result.
Heretofore, there is known the storage area network (SAN) that is a system for connecting storage devices through a network in contrast with a system for connecting computers through a network. In the SAN, the storage devices are usually connected through a fiber channel switch. Further, the storage devices can be connected to the storage devices installed in a remote place through a channel extender. The SAN is characterized in that transfer between the storage devices is made stably at a high speed.
Usually, a predetermined management system is provided for such storage devices and the management system can be used to make setting and management of various kinds to the storage devices. For example, it can be established that various use situations of the storage devices are displayed, which disk is used from which computer, what is the capacity thereof and so on.
On the other hand, JP-A-8-137725 discloses the technique of predicting the performance of a computer system upon change of an operation amount and modification of the hardware configuration.
The conventional management system of the storage devices can display information about various use situations in the devices. Accordingly, a user can be informed that the use capacity of a certain computer exceed 90% (the rate of the use capacity to the maximum capacity assigned to the computer) to thereby make treatment such as increase of the assignment amount or increase of a disk device.
In the conventional management system, however, since the displayed use situations are information representing the situations at the present time, it is impossible to understand how the situations are changed in the future. For example, even when the use capacity exceeds 90%, there is a case where any problem does not arise even if it is left as it is for a while when an increase rate is small, although it is necessary to make treatment such as increase of an assignment capacity promptly when the increase rate is large. The conventional management system cannot make management responsive to such change in the future.
Further, there is known the technique of predicting the performance as described in JP-A-8-137725, while even if the technique is applied to management of the storage devices, a prediction value is merely obtained and a manager is required to judge how treatment is made in response to the prediction value.
It is an object of the present invention to make it possible to analyze condition information of storage devices and manage the storage devices on the basis of the analyzed result.
In order to achieve the above object, according to the present invention, in order to manage the storage devices utilizing a storage area network (SAN), there is established policy that prescribes actions each executed when some object to be judged satisfies some condition with regard to storage devices. Information concerning use situations of the storage devices is collected and a value of an object to be judged of the policy is calculated on the basis of the collected information to judge whether the value of the object to be judged satisfies a condition of the policy or not. In this judgment, when the condition of the policy is satisfied, the action of the policy is executed. The objects to be judged of the policy include, for example, use capacity of storage, the number of I/O operations of disk or fragmentation. Further, the objects to be judged include actual result values in the past, current values or prediction values in the future relative thereto. Execution of the action of the policy is made by transmitting a command of the action to an integration control unit which integrates operations to each logical layer of the SAN to be executed.
An embodiment of the present invention is now described with reference to the accompanying drawings.
The server 101, 102 and 103 are connected to the switches 111 and 112 through lines 141 to 146 of SCSI. The switches 111 and 112 are connected to the switches 113 and 114 through fiber channels 151 to 154 and the switches 113 and 114 are connected to the disk devices 121, 122 and 123 through fiber channels 161 to 166. Further, these apparatuses are connected to a network shown by broken line 171. The protocol of the network 171 is TCP/IP. The TCP/IP network 171 is provided so as to control configuration and the like of the devices, detect trouble of the devices and so on. Commands for making control by Telnet are defined for the switches so as to make it possible to control the switches by the commands. In the embodiment, the TCP/IP network 171 is used to obtain the configuration information and the use situations of the storage network.
Agents 181, 182 and 182 which are programs for collecting information concerning the use situations of the storages are operated in the servers 101, 102 and 103. The agents 181, 182 and 183 collect information concerning, for example, use capacities of the storages used by the servers, the number of I/O operations of disk, fragmentation in each used area and the like. The collected information is transmitted to the performance manager 134 when a request is issued from the performance manager 134 by polling.
The performance manager 134 is operated in the performance management server 131. The performance manager 134 collects information of various kinds concerning the use situations of the storages in each server by polling the agents 181, 182 and 183 operating in the servers using the storages. Further, the performance manager analyzes the collected information. This analyzation contains transition or change in the past, current actual results and prediction of the future. For example, as prediction values of the use capacity of a certain disk, prediction values of the use capacity predicted after one week and after one month are calculated. What information is collected and what analyzation is made can be designated by the user's policy. This will be described in detail later.
The management server 132 manages the storage devices of the system. The management client 133 accesses to the management server 132 and instructs the management server to display the configuration information of the storage devices. The management client 133 displays a picture showing the configuration information of the storage network sent from the management server 132 in response to the instruction. Further, the management client 133 can access to the management server 132 to perform control of various kinds to the storage devices. For example, a maximum value of the use capacity for a certain server is set and changed. Moreover, the management client 133 can be connected to the management server 132 to select the policy. The policy is to relates to management of the storage devices. Particularly, when the prediction values of various kinds concerning the storage devices satisfy a predetermined condition, the system can establish the policy prescribing how the system cope therewith.
For example, the system can establish the policy that “if the prediction value of the use capacity of the disk device after one month for the server 1 exceeds 100%, data in the disk device is moved to another disk device”. This is the policy instructing that the prediction value of the use capacity of the disk device after one month for the disk device used by the server 1 is calculated and if it exceeds 100%, data in the disk device is automatically moved to another disk device. Such policy can be established to thereby manage the storage device on the basis of the prediction value in the future. Similarly, management can be also made on the basis of the actual results at the present time and in the past. A definite example of the policy will be described later.
Servers 207 and 208 are computers corresponding to the servers 101, 102, 103 and the like of
A performance manager 201 corresponds to the performance manager 134 operated in the performance management server 131 of
The performance manager 201 is connected to a trigger issuing rule memory unit 202, a repository memory unit 203 and a collected information and analyzed result memory unit 204. The collected information and analyzed result memory unit 204 stores information collected by means of the agents 261 and 271 and analyzed results obtained by analyzing the collected information. The trigger issuing rule memory unit 202 stores a trigger issuing rule. The trigger issuing rule is a rule representing a condition for issuing a trigger (condition for starting the policy) to the management server 205. When the collected information and the analyzed result satisfy the trigger issuing rule, the trigger is issued to the management server 205. The repository memory unit 203 stores repository. The repository includes configuration information of the network, configuration information of the storages, scenario definition information, schedule information, policy definition information, history information and the like.
The management server 205 is provided with a policy memory unit 206. The policy memory unit 206 stores the policy selected by the user. The management server 205 includes a policy selection unit 241, a condition judgment unit 242 and an action execution unit 243. Further, the performance manager 201 includes a trigger issuing unit 211, an analyzation unit 212 and a collection unit 213.
The policy selection unit 241 of the management server 205 receives policy selection operation by the user and stores the selected policy in the policy memory unit 206. The policy selection operation can be made from the management client (133 of
The collection unit 213 of the performance manager 201 polls the agents 261 and 271 operating in the servers to thereby collect information of various kinds in the servers. The collected results are stored in the collected information and analyzed result memory unit 204. The analyzation unit 212 analyzes the collected information and stores the analyzed result in the collected information and analyzed result memory unit 204. The trigger issuing unit 211 judges whether the analyzed result is coincident with the trigger issuing rule 202 or not. When the trigger issuing rule is satisfied, a trigger signal is issued to be transmitted to the management server 205.
The management server 205 receives the issued trigger and the condition judgment unit 242 judges the condition containing the trigger to judge whether action is executed or not. When the action is to be executed, the action execution unit 243 executes the predetermined action.
In execution of the action, an instruction that causes the computer, the storage device, the program or the like to execute processing may be transmitted or processing in execution may be interrupted or other processing may be executed.
The data of the trigger signal transmitted from the performance manager to the management server may contain information collected from the computer in which the program for managing the storage device is executed, information of the analyzed results of information of various kinds, information inherent in the performance manager and other information. For example, when a plurality of performance managers are operated and the management server receives information from the plurality of performance managers, an identification is assigned to each of the performance manager and the performance managers attach their own identification to data of the analyzed result to be transmitted to the management server to transmit the data to the management server so that the management server can identify the performance manager transmitting the data to thereby make processing.
In
In the above configuration, the policy for managing the storage devices can be executed on the basis of transition in the past, current actual results and prediction of the future.
In
Management tools 311 to 315 in the layers are provided. DBMS (DB management system) 311 is a tool for managing the databases and the files in the layers of DB 301 and File 302. A file system control unit 312 is a tool for managing the files and the file systems in the layers of File 302 and File System 303. A logical volume control unit 313 is a tool for managing the logical volumes in the layer of LVM 304 mainly and further makes even management of the layers of File System 303 and Device File 305 in relation to the logical volumes. RAID apparatus control unit 315 is a tool for managing the RAID in the layer of Device File 305 to LDEV 307.
The data protection and integration control unit 320 manages the correspondence relation of objects representing how objects correspond to in the layers from DB 301 of the higher rank to the LDEV 307 of the lower rank. Further, the data protection and integration control unit 320 provides the function for making it possible to execute operation such as backup and restore by one command. Consequently, the integration manager is merely required to give the data protection and integration control unit 320 an instruction by means of a simple command such as, for example, “back the table of DB up” even if the manager does not have any knowledge about all the layers. This reason is that the data protection and integration control unit 320 can follow the layers from the table name to the lower layer to easily understand the physical disk in which the table is stored and the group of the disk to be backed up is previously defined. Numeral 321 represents repository (management DB) corresponding to the repository memory unit 203 of
In the embodiment, the action execution unit 243 of
The objects to be judged are measured values of various kinds concerning the storages and contains transition in the past, current actual results and prediction of the future. As the objects to be judged, the capacity, the number of I/O operations of disk and fragmentation can be designated. The capacity contains, for example, “the current use capacity (the use capacity of the memory medium at the present time) of the server 1”, “the prediction value of the use capacity after one month of the server 1” and the like. Further, a value utilizing a differential value of the use capacity such as an increase rate of the use capacity calculated from transition of the use capacity in the past is allowed to be contained in the objects to be judged. With regard to the number of I/O operations of disk, “the number of I/O operations of disk during a predetermined period”, “an increase rate at the present time of the number of I/O operations of disk”, “a cumulative value of the numbers of I/O operations of disk during a predetermined period”, “a prediction value of the number of I/O operation of disk of tomorrow” and the like are allowed to be contained in the objects to be judged. With regard to the fragmentation, “fragmentation at the present time”, “a prediction value of fragmentation after one month” and the like are allowed to be contained in the objects to be judged. Numeral 401 represents an example of a list box for establishing the objects to be judged. Although not shown in
The conditions are established for the objects to be judged. For example, the condition that the object to be judged “exceeds 40%”, “exceeds 60%” and the like can be designated. Numeral 402 represents the list box for establishing the conditions for the objects to be judged.
The action is executed when the object to be judged satisfies the condition. Numeral 403 represents an action description area in which the action is described and a command of the action desired to be executed is described in this area. Describable commands are commands for the data protection and integration control unit 320 explained with reference to
to copy a predetermined file to another disk device;
to delete a temporary file;
to compress a predetermined file;
to back up to a tape or the like;
to issue a message for adding a disk;
to inform by means of an electronic mail or a picture display that the condition for the object to be judged is satisfied;
to move a file having a high use frequency to another file;
to disperse loads by duplication of paths; and
to move files so that files roughly resembling or equal to each other in the capacity thereof are stored in the same disk.
The list box is used to make selection by clicking a downward triangular mark at the right, although the method of establishing the policy is not limited thereto.
In the policy defining that (3) “when the prediction value of the use capacity of disk after two month for the server 1 exceeds 100%, the temporary file in the disk device is deleted and when the prediction value of the use capacity of disk after two month exceeds 100% even after the deletion, data stored in the disk device is compressed”, the object to be judged is “the prediction value of the use capacity of disk after two month for the server 1”, the condition is “to exceed 100%” and the action is “to delete the temporary file in the disk device used by the server 1 and when the prediction value of the use capacity of disk after two month even after the deletion, data in the disk device is deleted”. In this case, the action is to be described using the script. The data protection and integration control unit 320 has the function of interpreting the script and executing it.
In the policy defining that (4) “when the prediction value of the use capacity of disk after one month for the disk device 1 exceeds 80% and when the prediction value of the use capacity of disk after one month for the disk device 2 exceeds 80%, the message that a disk device should be added is displayed”, the object to be judged is “the prediction value of the use capacity of disk after one month for the disk device 1” and “the prediction value of the use capacity of disk after one month for the disk device 2”, the condition is “to exceed 80%” for both the objects and the action is “the message that a disk device should be added is displayed”. This is an example that sets of the objects to be judged and the conditions are combined by AND. In this case, a user interface for designating the combination of such conditions is required.
In the policy defining that (5) “when the increase rate at this moment in time of the number of I/O operations (for a predetermined period) for the disk device 1 is equal to or larger than a predetermined value RP, the file having the use frequency within a predetermined period in the past which is equal to or larger than XX times is moved to another disk device”, the object to be judged is “the increase rate at this moment in time of the number of I/O operations (for a predetermined period) for the disk device 1”, the condition is “to be equal to or larger than a predetermined value RP” and the action is “the file having the use frequency within a predetermined period in the past which is equal to or larger than XX times is moved to another disk device”.
In the policy defining that (6) “when the prediction value of the fragmentation after one month for the disk device 1 is equal to or larger than 50%, optimization is made”, the object to be judged is “the prediction value of the fragmentation after one month for the disk device 1”, the condition is “to be equal to or larger than 50%” and the action is “the disk device 1 is optimized”.
b) shows the procedure of information transmission processing. This information transmission processing is executed when polling is performed from the performance manager. First, in step 704, necessary information is obtained from the buffer 703. In step 705, the obtained information is transmitted to the performance manager 201. The information obtained to be transmitted is to be designated by polling.
Further, the trigger issuing rule 202 is composed of the objects to be judged and the conditions of the established policy. When the policy is established, the objects to be judged and the conditions of the policy are transmitted from the management server 206 to the performance manager 201 to be stored as the trigger issuing rule 202.
Further, the condition to be judged in step 902 is not only the condition relative to the trigger transmitted from the performance management manger 201. There is a case where the management server 205 makes judgment independently besides the judgment by the performance manager 201. For example, when the condition in a mere storage configuration which is not related to the prediction value relative to the storages is judged, the judgment can be made by only the management server 205. Further, a plurality of performance managers 201 may be provided in the system. For example, the storages may be classified into groups in accordance with any category and a plurality of performance manager may be provided as a performance manager given charge of a first group and a performance manager given charge of a second group and so on. In this case, when the policy includes a combination of AND and OR of triggers from the plurality of performance management mangers, whether the condition of the policy is effective or not is decided by the management server 205 (step 902).
In the embodiment, the performance management server and the management server are constructed separately, although they may be constructed on the same computer.
As described above, according to the embodiment, since the situation information of the storage devices constituting the SAN (Storage Area Network) is collected to calculate the actual result value in the past, the current value or the prediction value in the future relative to the use capacity of the storages, the number of I/O operations of disk or the fragmentation and the predetermined action is executed when the condition of the policy established thereby is satisfied, flexible storage management can be made in conformity with the situations. Particularly, since management is made on the basis of the prediction value in the future, treatment can be made in accordance with the situations predicted after one month, for example.
As described above, according to the present invention, since the situation information of the storage devices is collected and it is instructed that processing is executed on the basis of the collected information and previously defined contents of processing and execution, the storage devices can be managed flexibly in accordance with the state.
It should be further understood by those skilled in the art that although the foregoing description has been made on embodiments of the invention, the invention is not limited thereto and various changes and modifications may be made without departing from the spirit of the invention and the scope of the appended claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2002-148810 | May 2002 | JP | national |
This application is a continuation application of U.S. application Ser. No. 10/366,554, filed Feb. 14, 2003, now U.S. Pat. No. 7,246,105, the entirety of which is incorporated herein by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5379423 | Mutoh et al. | Jan 1995 | A |
| 5398337 | Ito | Mar 1995 | A |
| 5537585 | Blickenstaff et al. | Jul 1996 | A |
| 5619690 | Matsunami et al. | Apr 1997 | A |
| 5941972 | Hoese et al. | Aug 1999 | A |
| 5956750 | Yamamoto et al. | Sep 1999 | A |
| 5991753 | Wilde | Nov 1999 | A |
| 6041381 | Hoese | Mar 2000 | A |
| 6065087 | Keaveny et al. | May 2000 | A |
| 6098129 | Fukuzawa et al. | Aug 2000 | A |
| 6209023 | Dimitroff et al. | Mar 2001 | B1 |
| 6275898 | DeKoning | Aug 2001 | B1 |
| 6327614 | Asano et al. | Dec 2001 | B1 |
| 6446141 | Nolan et al. | Sep 2002 | B1 |
| 6510456 | Ikegaya et al. | Jan 2003 | B1 |
| 6598174 | Parks et al. | Jul 2003 | B1 |
| 6640278 | Nolan et al. | Oct 2003 | B1 |
| 6647474 | Vishlitzky et al. | Nov 2003 | B2 |
| 6654830 | Taylor et al. | Nov 2003 | B1 |
| 6745207 | Reuter et al. | Jun 2004 | B2 |
| 6751702 | Hsieh et al. | Jun 2004 | B1 |
| 6810462 | Matsunami et al. | Oct 2004 | B2 |
| 6922761 | O'Connell et al. | Jul 2005 | B2 |
| 6938039 | Bober et al. | Aug 2005 | B1 |
| 6947940 | Anderson et al. | Sep 2005 | B2 |
| 6950920 | Mizuno et al. | Sep 2005 | B1 |
| 6973455 | Vahalia et al. | Dec 2005 | B1 |
| 20010054133 | Murotani et al. | Dec 2001 | A1 |
| 20020059539 | Anderson | May 2002 | A1 |
| 20020062387 | Yatziv | May 2002 | A1 |
| 20020161855 | Manczak et al. | Oct 2002 | A1 |
| 20030046270 | Leung et al. | Mar 2003 | A1 |
| 20030061440 | Elliott | Mar 2003 | A1 |
| 20030065873 | Collins et al. | Apr 2003 | A1 |
| 20030074523 | Johnson | Apr 2003 | A1 |
| 20030112538 | Smith | Jun 2003 | A1 |
| 20030139900 | Robison | Jul 2003 | A1 |
| 20030177261 | Sekiguchi et al. | Sep 2003 | A1 |
| 20030182288 | O'Connell et al. | Sep 2003 | A1 |
| 20030182525 | O'Connell et al. | Sep 2003 | A1 |
| 20040044854 | Gibble et al. | Mar 2004 | A1 |
| 20040068561 | Yamamoto et al. | Apr 2004 | A1 |
| 20040083202 | Mu et al. | Apr 2004 | A1 |
| 20040098394 | Merritt et al. | May 2004 | A1 |
| 20040107315 | Watanabe et al. | Jun 2004 | A1 |
| 20040143563 | Saake et al. | Jul 2004 | A1 |
| 20040143648 | Koning et al. | Jul 2004 | A1 |
| 20040162940 | Yagisawa et al. | Aug 2004 | A1 |
| 20040199515 | Penny et al. | Oct 2004 | A1 |
| 20040210724 | Koning et al. | Oct 2004 | A1 |
| 20040260862 | Anderson | Dec 2004 | A1 |
| 20050097126 | Cabrera et al. | May 2005 | A1 |
| 20050120189 | Black | Jun 2005 | A1 |
| 20050149528 | Anderson et al. | Jul 2005 | A1 |
| 20050149671 | Suzuki et al. | Jul 2005 | A1 |
| 20050172097 | Voigt et al. | Aug 2005 | A1 |
| Number | Date | Country |
|---|---|---|
| 3-223941 | Oct 1991 | JP |
| H5-307532 | Nov 1993 | JP |
| H7-69927 | Jul 1995 | JP |
| 8-137725 | May 1996 | JP |
| 09-259037 | Oct 1997 | JP |
| 09274544 | Oct 1997 | JP |
| 09297699 | Nov 1997 | JP |
| 10-301720 | Nov 1998 | JP |
| 2002-229740 | Aug 2002 | JP |
| WO 02069159 | Sep 2002 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20070124273 A1 | May 2007 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10366554 | Feb 2003 | US |
| Child | 11648654 | US |
