[ 
https://issues.apache.org/jira/browse/HDFS-13762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16773040#comment-16773040
 ] 

Sammi Chen commented on HDFS-13762:
-----------------------------------

Thanks [~PhiloHe] for continue working on this.  Here are some feebacks for 
008.patch.
 1. NativeIO.java,  suggest to define the different PMDK support code and it's 
meaning using a enum, so it will be easy to map the code to description.
 2. NativeIO.c, it would be nice to refactor the error message, to avoid the 
potential array overflow if pmem_errormsg() returns too long content. There are 
several piece of similar code
   
{code:java}
char msg[1000];
   snprintf(msg, sizeof(msg), "Fail to unmap region. address: %x, length: %x, 
error msg: %s", address, length, pmem_errormsg());
{code}

3. Support persistent memeory cache on Windows can be a follow-up task since 
Windows is not very common used in user environment. For build patch on 
Windows, if it's a goal of this JIRA, please update the "Building on Windows" 
in BUILDING.txt for any specific steps user need to take. If it's not a goal of 
this JIRA, make sure build on Windows will have clear message for this case. 
4. Cmakelists.txt  "#Require ISA-L" is for ISA-L, better keep it. 
5. make PmemVolumeManager pmemManager final

6. 
 This piece of code has the chance to delete user files accidentally. Better to 
provide user a choice to decide auto delete or not. 
{code:java}
   // Remove all files under the volume. Files may been left after a
          // unexpected data node restart.
          FileUtils.cleanDirectory(locFile);
{code}

7. make count in PmemVolumeManager final
8. fillBuffer in MemoryMappedBlock and PmemMappedBlock are the same. refactor 
to keep only one piece of code 





> Support non-volatile storage class memory(SCM) in HDFS cache directives
> -----------------------------------------------------------------------
>
>                 Key: HDFS-13762
>                 URL: https://issues.apache.org/jira/browse/HDFS-13762
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>          Components: caching, datanode
>            Reporter: Sammi Chen
>            Assignee: Feilong He
>            Priority: Major
>         Attachments: HDFS-13762.000.patch, HDFS-13762.001.patch, 
> HDFS-13762.002.patch, HDFS-13762.003.patch, HDFS-13762.004.patch, 
> HDFS-13762.005.patch, HDFS-13762.006.patch, HDFS-13762.007.patch, 
> HDFS-13762.008.patch, SCMCacheDesign-2018-11-08.pdf, SCMCacheTestPlan.pdf
>
>
> No-volatile storage class memory is a type of memory that can keep the data 
> content after power failure or between the power cycle. Non-volatile storage 
> class memory device usually has near access speed as memory DIMM while has 
> lower cost than memory.  So today It is usually used as a supplement to 
> memory to hold long tern persistent data, such as data in cache. 
> Currently in HDFS, we have OS page cache backed read only cache and RAMDISK 
> based lazy write cache.  Non-volatile memory suits for both these functions. 
> This Jira aims to enable storage class memory first in read cache. Although 
> storage class memory has non-volatile characteristics, to keep the same 
> behavior as current read only cache, we don't use its persistent 
> characteristics currently.  
>  
>  
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org

Reply via email to