accumulo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Moss <>
Subject Re: BatchWriter Improvements - An end user's perspective
Date Tue, 30 Aug 2016 18:20:42 GMT
Appreciate your input, Keith. Some inlining below.

On 2016-08-29 16:24 (-0400), Keith Turner <> wrote:
> On Fri, Aug 26, 2016 at 12:21 PM, Michael Moss <> wrote:>
> > Hello, Folks.>
> >>
> > As I look at the following tickets, I thought it might be useful to
> > how we are using the BatchWriter, some of the challenges we've had,
> > thoughts about it's redesign and how we might get involved.>
> >>
> >>
> >>
> >>
> >>
> > One of our primary use cases of the BatchWriter is from within a Storm>
> > topology, reading from Kafka. Generally speaking, storm might be
> > a a single or small set of mutations at a time (low latency), or in
> > batches with Trident (higher throughput). In addition to ACCUMULO-2990
> > TimedOutException, which then throws MutationsRejectedException and>
> > requires a new connection to be made), one of our requirements is to
> > that any given thread's mutations are the ones which are flushed and
> > others (pseudo transactions). Otherwise, we might get a failure for a>
> > mutation which belongs to another thread (and already ACKed by Storm)
> > means we don't have a 'handle' on that offset anymore in Kafka to
> > the failure - i.e. the message could be 'lost'.>
> >>
> > Despite being threadsafe, we end up using a single BatchWriter per
> How many threads are you having each batch writer use?>

We've tuned it to be anywhere between the default of 3, to 1-x * # of cpus.
The latest issue we ran into was under high write frequency, the queries
requesting metadata to locate tablets were being queued up on the server
and that was causing timeouts (in an endless loop - timeout, new
batchwriter, metadata query, timeout...). Increasing batching (decreasing
write frequency) helps with that, as well as increasing timeouts. I don't
know this code well yet, but it looks like there is
one TabletLocatorImpl.metaCache per TabletServerBatchWriter. I'm wondering
if maybe that metadata could instead be shared across all BatchWriters on
the Connector instead?

> > to make reasoning about the above simpler, but this creates a resource>
> > issue - number of connections to accumulo and zk.>
> >>
> > This all makes me wonder what the design goals might have been for the>
> > current version of the driver and if the efforts to rewrite it might>
> > benefit from incorporating elements to address some of these use cases>
> > above.>
> >>
> > What can we learn from how drivers for other "NoSQL" databases are>
> > implemented? Would it make sense to remove all the global variables>
> > ("somethingFailed"), thread sleep/notify, frequent calls to>
> > "checkForFailures()" and consider using a 'connection pool' model
> > writes are single-threaded, linearized and isolated during the
> > lease? Could we make the client non-blocking and with optional
> > so multiple writes could share a connection and allow interleaving of>
> > operations (with individual acks)?>
> I am hoping to work on async batch writer, conditional writter, and>
> scanners at some point.   I think this may address some of the things>
> you brought up.   If I do start working on it, I will seek input as>
> early as possible.  For example if we had something that accepted a>
> batch of mutation to write and returned a>
> CompletableFuture<BatchResult>, then BatchResult could relay failure>
> information only related to that batch.  Also, flush() may not be>
> needed since a user could just wait on the future if desired.>
> I'm hoping when something like this exist in the API, optimizations>
> like pipelining could be made later if not done when its added.>
> >>
> > Looking forward to hearing everyone's thoughts.>
> >>
> > -Mike>

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message