Method, apparatus and device for deleting distributed system file, and storage medium
View Patent ↗A method, apparatus and device for deleting a distributed system file, and a storage medium, comprising: querying whether an incomplete file deletion operation exists under a sub-tree root corresponding to a certain MDS (S 102 ); if there is an incomplete file deletion operation, continuing to delete data under the sub-tree root corresponding to the MDS, and determining whether the sub-tree root is a copy (S 103 ); and if the sub-tree root is a copy, deleting the sub-tree root copy in a memory of the MDS (S 104 ). By means of the steps, the number of copies of the sub-tree root in the distributed file system can be reduced, thereby reducing the number of interaction times of master copy locking between different MDSs, improving the file deletion efficiency, and improving user friendliness and differentiated competitiveness of a product.
1. A method for deleting a file in a distributed system, characterized by comprising:
querying whether an incomplete file deletion operation exists under a sub-tree root corresponding to a certain metadata server (MDS);
continuing to delete data under the sub-tree root corresponding to the MDS if there is an incomplete file deletion operation, and determining whether the sub-tree root is a copy; and
deleting the sub-tree root copy in a memory of the MDS if the sub-tree root is a copy;
creating a timed deleting task in the distributed system if there is no incomplete file deletion operation; and
balancing the number of sub-tree roots to be deleted among different MDSs according to the created timed deleting task;
wherein the step of balancing the number of sub-tree roots to be deleted among different MDSs according to the created timed deleting task specifically comprises:
acquiring and recording the number of sub-tree roots to be deleted with the timed deleting task on each MDS; and
calculating an average value of the number of sub-tree roots to be deleted on all MDSs; and
migrating sub-tree roots to be deleted to the MDSs having a number of sub-tree roots to be deleted lower than the average value sequentially from the MDS having the largest number of sub-tree roots to be deleted, until no MDS having a number of sub-tree roots to be deleted less than the average value exists.
2. The method for deleting a file in a distributed system according to claim 1 , characterized by further comprising:
performing segmentation and deletion operations on a file to be deleted according to the actual size of the file.
3. The method for deleting a distributed system file according to claim 2 , characterized in that performing segmentation and deletion operations on a file to be deleted according to the actual size of the file specifically comprises:
segmenting the file to be deleted into a plurality of objects to be deleted of which the size is equal to a stripe value according to the actual size of the file;
sending the objects to be deleted to an OSD corresponding to the objects to be deleted sequentially according to a preset sequence; and
performing a disk deletion operation on all the objects to be deleted by means of the OSD.
4. A device for deleting a file in a distributed system, characterized by comprising a processor and a memory, wherein the memory has a computer program stored therein, and the computer program, when executed by the processor, causes the processor to perform the following steps:
querying whether an incomplete file deletion operation exists under a sub-tree root corresponding to a certain metadata server (MDS);
continuing to delete data under the sub-tree root corresponding to the MDS if there is an incomplete file deletion operation, and determining whether the sub-tree root is a copy; and
deleting the sub-tree root copy in a memory of the MDS if the sub-tree root is a copy;
creating a timed deleting task in the distributed system if there is no incomplete file deletion operation; and
balancing the number of sub-tree roots to be deleted among different MDSs according to the created timed deleting task;
wherein the step of balancing the number of sub-tree roots to be deleted among different MDSs according to the created timed deleting task specifically comprises:
acquiring and recording the number of sub-tree roots to be deleted with the timed deleting task on each MDS; and
calculating an average value of the number of sub-tree roots to be deleted on all MDSs; and
migrating sub-tree roots to be deleted to the MDSs having a number of sub-tree roots to be deleted lower than the average value sequentially from the MDS having the largest number of sub-tree roots to be deleted, until no MDS having a number of sub-tree roots to be deleted less than the average value exists.
5. A computer readable storage medium, having a computer program stored therein, characterized in that the computer program, when executed by a processor, causes the processor to perform the following steps:
querying whether an incomplete file deletion operation exists under a sub-tree root corresponding to a certain metadata server (MDS);
continuing to delete data under the sub-tree root corresponding to the MDS if there is an incomplete file deletion operation, and determining whether the sub-tree root is a copy; and
deleting the sub-tree root copy in a memory of the MDS if the sub-tree root is a copy;
creating a timed deleting task in the distributed system if there is no incomplete file deletion operation; and
balancing the number of sub-tree roots to be deleted among different MDSs according to the created timed deleting task;
wherein the step of balancing the number of sub-tree roots to be deleted among different MDSs according to the created timed deleting task specifically comprises:
acquiring and recording the number of sub-tree roots to be deleted with the timed deleting task on each MDS; and
calculating an average value of the number of sub-tree roots to be deleted on all MDSs; and
migrating sub-tree roots to be deleted to the MDSs having a number of sub-tree roots to be deleted lower than the average value sequentially from the MDS having the largest number of sub-tree roots to be deleted, until no MDS having a number of sub-tree roots to be deleted less than the average value exists.