accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Adam Fuchs (JIRA)" <>
Subject [jira] [Commented] (ACCUMULO-4669) RFile can create very large blocks when key statistics are not uniform
Date Fri, 30 Jun 2017 19:13:00 GMT


Adam Fuchs commented on ACCUMULO-4669:

Interesting that the case that caused these large blocks for us also had to do with long URLs.
However, we weren't hash distributing them in this case, so they ended up bunching up and
causing a localized deviation in key size statistics. How long is the hash you were using?

We should really think about using something closer to a trie structure for files, where the
index tree is the prefix and the data blocks only contain tails of keys. RFiles still have
a lot of redundancy in the indexes and between relative-key-encoded blocks, especially when
using large key/values and not many in a single block.

> RFile can create very large blocks when key statistics are not uniform
> ----------------------------------------------------------------------
>                 Key: ACCUMULO-4669
>                 URL:
>             Project: Accumulo
>          Issue Type: Bug
>          Components: core
>    Affects Versions: 1.7.2, 1.7.3, 1.8.0, 1.8.1
>            Reporter: Adam Fuchs
>            Assignee: Keith Turner
>            Priority: Blocker
>             Fix For: 1.7.4, 1.8.2, 2.0.0
> RFile.Writer.append checks for giant keys and avoid writing them as index blocks. This
check is flawed and can result in multi-GB blocks. In our case, a 20GB compressed RFile had
one block with over 2GB raw size. This happened because the key size statistics changed after
some point in the file. The code in question follows:
> {code}
>     private boolean isGiantKey(Key k) {
>       // consider a key thats more than 3 standard deviations from previously seen key
sizes as giant
>       return k.getSize() > keyLenStats.getMean() + keyLenStats.getStandardDeviation()
* 3;
>     }
> ...
>       if (blockWriter == null) {
>         blockWriter = fileWriter.prepareDataBlock();
>       } else if (blockWriter.getRawSize() > blockSize) {
>         ...
>         if ((prevKey.getSize() <= avergageKeySize || blockWriter.getRawSize() >
maxBlockSize) && !isGiantKey(prevKey)) {
>           closeBlock(prevKey, false);
> ...
> {code}
> Before closing a block that has grown beyond the target block size we check to see that
the key is below average in size or that the block is 1.1 times the target block size (maxBlockSize),
and we check that the key isn't a "giant" key, or more than 3 standard deviations from the
mean of keys seen so far.
> Our RFiles often have one row of data with different column families representing various
forward and inverted indexes. This is a table design similar to the WikiSearch example. The
first column family in this case had very uniform, relatively small key sizes. This first
column family comprised gigabytes of data, split up into roughly 100KB blocks. When we switched
to the next column family the keys grew in size, but were still under about 100 bytes. The
statistics of the first column family had firmly established a smaller mean and tiny standard
deviation (approximately 0), and it took over 2GB of larger keys to bring the standard deviation
up enough so that keys were no longer considered "giant" and the block could be closed.
> Now that we're aware, we see large blocks (more than 10x the target block size) in almost
every RFile we write. This only became a glaring problem when we got OOM exceptions trying
to decompress the block, but it also shows up in a number of subtle performance problems,
like high variance in latencies for looking up particular keys.
> The fix for this should produce bounded RFile block sizes, limited to the greater of
2x the maximum key/value size in the block and some configurable threshold, such as 1.1 times
the compressed block size. We need a firm cap to be able to reason about memory usage in various
> The following code produces arbitrarily large RFile blocks:
> {code}
>   FileSKVWriter writer = RFileOperations.getInstance().openWriter(filename, fs, conf,
>   writer.startDefaultLocalityGroup();
>   SummaryStatistics keyLenStats = new SummaryStatistics();
>   Random r = new Random();
>   byte [] buffer = new byte[minRowSize]; 
>   for(int i = 0; i < 100000; i++) {
>     byte [] valBytes = new byte[valLength];
>     r.nextBytes(valBytes);
>     r.nextBytes(buffer);
>     ByteBuffer.wrap(buffer).putInt(i);
>     Key k = new Key(buffer, 0, buffer.length, emptyBytes, 0, 0, emptyBytes, 0, 0, emptyBytes,
0, 0, 0);
>     Value v = new Value(valBytes);
>     writer.append(k, v);
>     keyLenStats.addValue(k.getSize());
>     int newBufferSize = Math.max(buffer.length, (int) Math.ceil(keyLenStats.getMean()
+ keyLenStats.getStandardDeviation() * 4 + 0.0001));
>     buffer = new byte[newBufferSize];
>     if(keyLenStats.getSum() > targetSize)
>       break;
>   }
>       writer.close();
> {code}
> One telltale symptom of this bug is an OutOfMemoryException thrown from a readahead thread
with message "Requested array size exceeds VM limit". This will only happen if the block cache
size is big enough to hold the expected raw block size, 2GB in our case. This message is rare,
and really only happens when allocating an array of size Integer.MAX_VALUE or Integer.MAX_VALUE-1
on the hotspot JVM. Integer.MAX_VALUE happens in this case due to some strange handling of
raw block sizes in the BCFile code. Most OutOfMemoryExceptions have different messages.

This message was sent by Atlassian JIRA

View raw message