cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Björn Hegerfors (JIRA) <>
Subject [jira] [Commented] (CASSANDRA-9666) Provide an alternative to DTCS
Date Tue, 29 Mar 2016 16:03:25 GMT


Björn Hegerfors commented on CASSANDRA-9666:

So, I also think that the tiering should have some value, though I lack concrete benchmarks
to show it.

That said, the problems that DTCS still suffers from with the latest patches, according to
what's being said here, would perhaps be best attacked by a breaking change. So I see an opportunity
to do something about it, whether the name stays or changes to TWCS (I kind of like that name).

For example, what's holding back CASSANDRA-11056 is a fear that the transition won't be completely
smooth. The names of the options in DTCS is something that I would like to change towards
what TWCS has. max_window_size has become more central than base_time_seconds. Renaming max_window_size
to window_size and base_time_seconds to min_window_size (or even more obscure, or even removing
the option) should be better. If we require user interaction anyway, then applying CASSANDRA-11056
while we're at it would be easier.

Also, the idea in CASSANDRA-9013 and CASSANDRA-11407 and the recent recent dev@ mail thread,
shows another way of doing the tiering that would arguably be easier to understand. Combine
these suggestions, and we end up expressing the windowing forwards in time, rather than backwards.
DTCS was always about "what's the size of the smallest window", "how many of those make a
larger window" and "how long do we wait until we stop compacting a value". Instead, we would
get the TWCS way, which people seem to like more, where the main question is "how large are
the windows that we put the values into". And beyond that there's this ramp down in window
sizes just towards the end to accommodate for very write heavy workloads, but options controlling
that part a more of a detail, like the min_sstable_size option of STCS.

Is something like this what people really want? I.e. TWCS, but with tiering near the end of
the timeline, that users don't need to know or care about?

> Provide an alternative to DTCS
> ------------------------------
>                 Key: CASSANDRA-9666
>                 URL:
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Jeff Jirsa
>            Assignee: Jeff Jirsa
>             Fix For: 2.1.x, 2.2.x
>         Attachments: dtcs-twcs-io.png, dtcs-twcs-load.png
> DTCS is great for time series data, but it comes with caveats that make it difficult
to use in production (typical operator behaviors such as bootstrap, removenode, and repair
have MAJOR caveats as they relate to max_sstable_age_days, and hints/read repair break the
selection algorithm).
> I'm proposing an alternative, TimeWindowCompactionStrategy, that sacrifices the tiered
nature of DTCS in order to address some of DTCS' operational shortcomings. I believe it is
necessary to propose an alternative rather than simply adjusting DTCS, because it fundamentally
removes the tiered nature in order to remove the parameter max_sstable_age_days - the result
is very very different, even if it is heavily inspired by DTCS. 
> Specifically, rather than creating a number of windows of ever increasing sizes, this
strategy allows an operator to choose the window size, compact with STCS within the first
window of that size, and aggressive compact down to a single sstable once that window is no
longer current. The window size is a combination of unit (minutes, hours, days) and size (1,
etc), such that an operator can expect all data using a block of that size to be compacted
together (that is, if your unit is hours, and size is 6, you will create roughly 4 sstables
per day, each one containing roughly 6 hours of data). 
> The result addresses a number of the problems with DateTieredCompactionStrategy:
> - At the present time, DTCS’s first window is compacted using an unusual selection
criteria, which prefers files with earlier timestamps, but ignores sizes. In TimeWindowCompactionStrategy,
the first window data will be compacted with the well tested, fast, reliable STCS. All STCS
options can be passed to TimeWindowCompactionStrategy to configure the first window’s compaction
> - HintedHandoff may put old data in new sstables, but it will have little impact other
than slightly reduced efficiency (sstables will cover a wider range, but the old timestamps
will not impact sstable selection criteria during compaction)
> - ReadRepair may put old data in new sstables, but it will have little impact other than
slightly reduced efficiency (sstables will cover a wider range, but the old timestamps will
not impact sstable selection criteria during compaction)
> - Small, old sstables resulting from streams of any kind will be swiftly and aggressively
compacted with the other sstables matching their similar maxTimestamp, without causing sstables
in neighboring windows to grow in size.
> - The configuration options are explicit and straightforward - the tuning parameters
leave little room for error. The window is set in common, easily understandable terms such
as “12 hours”, “1 Day”, “30 days”. The minute/hour/day options are granular enough
for users keeping data for hours, and users keeping data for years. 
> - There is no explicitly configurable max sstable age, though sstables will naturally
stop compacting once new data is written in that window. 
> - Streaming operations can create sstables with old timestamps, and they'll naturally
be joined together with sstables in the same time bucket. This is true for bootstrap/repair/sstableloader/removenode.

> - It remains true that if old data and new data is written into the memtable at the same
time, the resulting sstables will be treated as if they were new sstables, however, that no
longer negatively impacts the compaction strategy’s selection criteria for older windows.

> Patch provided for : 
> - 2.1: 
> - 2.2:
> - trunk (post-8099): 
> Rebased, force-pushed July 18, with bug fixes for estimated pending compactions and potential
starvation if more than min_threshold tables existed in current window but STCS did not consider
them viable candidates
> Rebased, force-pushed Aug 20 to bring in relevant logic from CASSANDRA-9882

This message was sent by Atlassian JIRA

View raw message