[
https://issues.apache.org/jira/browse/HBASE-23887?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17285044#comment-17285044
]
Viraj Jasani commented on HBASE-23887:
--------------------------------------
Although this is not a pure bug fix as such and hence not a strong candidate
for patch release as per guidelines but on a broad level, I thought of these
factors to make it land on branch-2.4:
# Optional feature with simple config change
# Default L1 cache behaviour not changed
# No API changes involved
# 2.4.2 will likely come sooner than 2.5.0, and hence folks can get to try
this out with 2.4.2 and see how many workloads are best suited for AdaptiveLRU
(although latest claims on this Jira after last patch modification is that this
performs well with majority workflows, I know it's bit tough to find direct
comments/claims due to so much content present here)
[~ndimiduk] [~apurtell] If you think reason#4 above is not valid, I don't have
strong point to keep this on branch-2.4 and will revert it once you confirm.
{quote}It's a nice feature but not strongly compelling in that way in my
opinion. I do not object to its inclusion, for what its worth.
{quote}
[~apurtell] I agree, it's not too compelling to start release work for 2.5.0
and that's the reason why I wanted to include it on branch-2.4 to get it out as
part of 2.4.2 soon, but I totally understand that this cannot be the only
reason to roll it out in a patch release. Interested folks can backport these
commits in their internal branches to give it a shot. Hence, I am fine either
ways, no strong reason to keep it on branch-2.4.
> New L1 cache : AdaptiveLRU
> --------------------------
>
> Key: HBASE-23887
> URL: https://issues.apache.org/jira/browse/HBASE-23887
> Project: HBase
> Issue Type: Improvement
> Components: BlockCache, Performance
> Reporter: Danil Lipovoy
> Assignee: Danil Lipovoy
> Priority: Major
> Fix For: 3.0.0-alpha-1, 2.5.0, 2.4.2
>
> Attachments: 1582787018434_rs_metrics.jpg,
> 1582801838065_rs_metrics_new.png, BC_LongRun.png,
> BlockCacheEvictionProcess.gif, BlockCacheEvictionProcess.gif, PR#1257.diff,
> cmp.png, evict_BC100_vs_BC23.png, eviction_100p.png, eviction_100p.png,
> eviction_100p.png, gc_100p.png, graph.png, image-2020-06-07-08-11-11-929.png,
> image-2020-06-07-08-19-00-922.png, image-2020-06-07-12-07-24-903.png,
> image-2020-06-07-12-07-30-307.png, image-2020-06-08-17-38-45-159.png,
> image-2020-06-08-17-38-52-579.png, image-2020-06-08-18-35-48-366.png,
> image-2020-06-14-20-51-11-905.png, image-2020-06-22-05-57-45-578.png,
> image-2020-09-23-09-48-59-714.png, image-2020-09-23-10-06-11-189.png,
> ratio.png, ratio2.png, read_requests_100pBC_vs_23pBC.png, requests_100p.png,
> requests_100p.png, requests_new2_100p.png, requests_new_100p.png, scan.png,
> scan_and_gets.png, scan_and_gets2.png, wave.png, ycsb_logs.zip
>
>
> Hi!
> I first time here, correct me please if something wrong.
> All latest information is here:
> [https://docs.google.com/document/d/1X8jVnK_3lp9ibpX6lnISf_He-6xrHZL0jQQ7hoTV0-g/edit?usp=sharing]
> I want propose how to improve performance when data in HFiles much more than
> BlockChache (usual story in BigData). The idea - caching only part of DATA
> blocks. It is good becouse LruBlockCache starts to work and save huge amount
> of GC.
> Sometimes we have more data than can fit into BlockCache and it is cause a
> high rate of evictions. In this case we can skip cache a block N and insted
> cache the N+1th block. Anyway we would evict N block quite soon and that why
> that skipping good for performance.
> ---
> Some information below isn't actual
> ---
>
>
> Example:
> Imagine we have little cache, just can fit only 1 block and we are trying to
> read 3 blocks with offsets:
> 124
> 198
> 223
> Current way - we put the block 124, then put 198, evict 124, put 223, evict
> 198. A lot of work (5 actions).
> With the feature - last few digits evenly distributed from 0 to 99. When we
> divide by modulus we got:
> 124 -> 24
> 198 -> 98
> 223 -> 23
> It helps to sort them. Some part, for example below 50 (if we set
> *hbase.lru.cache.data.block.percent* = 50) go into the cache. And skip
> others. It means we will not try to handle the block 198 and save CPU for
> other job. In the result - we put block 124, then put 223, evict 124 (3
> actions).
> See the picture in attachment with test below. Requests per second is higher,
> GC is lower.
>
> The key point of the code:
> Added the parameter: *hbase.lru.cache.data.block.percent* which by default =
> 100
>
> But if we set it 1-99, then will work the next logic:
>
>
> {code:java}
> public void cacheBlock(BlockCacheKey cacheKey, Cacheable buf, boolean
> inMemory) {
> if (cacheDataBlockPercent != 100 && buf.getBlockType().isData())
> if (cacheKey.getOffset() % 100 >= cacheDataBlockPercent)
> return;
> ...
> // the same code as usual
> }
> {code}
>
> Other parameters help to control when this logic will be enabled. It means it
> will work only while heavy reading going on.
> hbase.lru.cache.heavy.eviction.count.limit - set how many times have to run
> eviction process that start to avoid of putting data to BlockCache
> hbase.lru.cache.heavy.eviction.bytes.size.limit - set how many bytes have to
> evicted each time that start to avoid of putting data to BlockCache
> By default: if 10 times (100 secunds) evicted more than 10 MB (each time)
> then we start to skip 50% of data blocks.
> When heavy evitions process end then new logic off and will put into
> BlockCache all blocks again.
>
> Descriptions of the test:
> 4 nodes E5-2698 v4 @ 2.20GHz, 700 Gb Mem.
> 4 RegionServers
> 4 tables by 64 regions by 1.88 Gb data in each = 600 Gb total (only FAST_DIFF)
> Total BlockCache Size = 48 Gb (8 % of data in HFiles)
> Random read in 20 threads
>
> I am going to make Pull Request, hope it is right way to make some
> contribution in this cool product.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
