Do you have the SSTables per read metric for before and after you increased
the key cache size? If it was high before, that may have been the culprit
meaning compaction tuning is in order.
On Fri, Aug 25, 2023, 12:35 PM Shaurya Gupta wrote:
> Thanks everyone.
> Updating this thread -
> We increa
Thanks everyone.
Updating this thread -
We increased the key cache size from 100 MB to 200 MB and we believe that
has brought down the latency from 40 ms p95 to 6 ms p95. I think there is
still scope for improvement as both writes and reads are presently at p95 6
ms. I would expect writes to be low
1. Check for Nagle/delayed-ack, but probably nodelay is getting set by the
driver so it shouldn't be a problem.
2. Check for network latency (just regular old ping among hosts, during
traffic)
3. Check your GC metrics and see if garbage collections line up with
outliers. Some tuning can help th
> The queries are rightly designed
Data modeling in Cassandra is 100% gray space; there unfortunately is no right
or wrong design. You'll need to share basic shapes / contours of your data
model for other folks to help you; seemingly innocuous things in a data model
can cause unexpected issues w
You’re going to have to help us help you 4.0 is pretty widely deployed. I’m not aware of a perf regression Can you give us a schema (anonymized) and queries and show us a trace ? On Aug 10, 2023, at 10:18 PM, Shaurya Gupta wrote:The queries are rightly designed as I already explained. 40 ms is wa
I would expect single digit ms latency on reads and writes. However, we have
not done any performance testing on Apache Cassandra 4.x.
Sean R. Durity
INTERNAL USE
From: Shaurya Gupta
Sent: Friday, August 11, 2023 1:16 AM
To: user@cassandra.apache.org
Subject: [EXTERNAL] Re: Cassandra p95
The queries are rightly designed as I already explained. 40 ms is way too
high as compared to what I seen with other DBs and many a times with
Cassandra 3.x versions.
CPU consumed as I mentioned is not high, it is around 20%.
On Thu, Aug 10, 2023 at 5:14 PM MyWorld wrote:
> Hi,
> P95 should not
40ms is definitely higher than expected. Have you run your queries with TRACING enabled to see where the latency is coming from?https://docs.datastax.com/en/cql-oss/3.3/cql/cql_reference/cqlshTracing.html40ms is also a fairly specific duration: https://eklitzke.org/the-caveats-of-tcp-nodelay> On Li
Hi,
P95 should not be a problem if rightly designed. Levelled compaction
strategy further reduces this, however it consume some resources. For read,
caching is also helpful.
Can you check your cpu iowait as it could be the reason for delay
Regards,
Ashish
On Fri, 11 Aug, 2023, 04:58 Shaurya Gupta