sia.hackernoon.com

Intro

In this article, we’ll observe another test added to db-benchmarks - 10+ million standard HTTP logs collected by Nginx on the e-commerce website zanbil.ir.

Data collection

We found the data collection on https://www.kaggle.com/datasets/eliasdabbas/web-server-access-logs and found it very interesting to test since the dataset represents a very standard Nginx HTTP access log. Here’s an example:

54.36.149.41 - - [22/Jan/2019:03:56:14 +0330] "GET /filter/27|13%20%D9%85%DA%AF%D8%A7%D9%BE%DB%8C%DA%A9%D8%B3%D9%84,27|%DA%A9%D9%85%D8%AA%D8%B1%20%D8%A7%D8%B2%205%20%D9%85%DA%AF%D8%A7%D9%BE%DB%8C%DA%A9%D8%B3%D9%84,p53 HTTP/1.1" 200 30577 "-" "Mozilla/5.0 (compatible; AhrefsBot/6.1; +http://ahrefs.com/robot/)" "-"
31.56.96.51 - - [22/Jan/2019:03:56:16 +0330] "GET /image/60844/productModel/200x200 HTTP/1.1" 200 5667 "https://www.zanbil.ir/m/filter/b113" "Mozilla/5.0 (Linux; Android 6.0; ALE-L21 Build/HuaweiALE-L21) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/66.0.3359.158 Mobile Safari/537.36" "-"
31.56.96.51 - - [22/Jan/2019:03:56:16 +0330] "GET /image/61474/productModel/200x200 HTTP/1.1" 200 5379 "https://www.zanbil.ir/m/filter/b113" "Mozilla/5.0 (Linux; Android 6.0; ALE-L21 Build/HuaweiALE-L21) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/66.0.3359.158 Mobile Safari/537.36" "-"
40.77.167.129 - - [22/Jan/2019:03:56:17 +0330] "GET /image/14925/productModel/100x100 HTTP/1.1" 200 1696 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"
91.99.72.15 - - [22/Jan/2019:03:56:17 +0330] "GET /product/31893/62100/%D8%B3%D8%B4%D9%88%D8%A7%D8%B1-%D8%AE%D8%A7%D9%86%DA%AF%DB%8C-%D9%BE%D8%B1%D9%86%D8%B3%D9%84%DB%8C-%D9%85%D8%AF%D9%84-PR257AT HTTP/1.1" 200 41483 "-" "Mozilla/5.0 (Windows NT 6.2; Win64; x64; rv:16.0)Gecko/16.0 Firefox/16.0" "-"
40.77.167.129 - - [22/Jan/2019:03:56:17 +0330] "GET /image/23488/productModel/150x150 HTTP/1.1" 200 2654 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"
40.77.167.129 - - [22/Jan/2019:03:56:18 +0330] "GET /image/45437/productModel/150x150 HTTP/1.1" 200 3688 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"
40.77.167.129 - - [22/Jan/2019:03:56:18 +0330] "GET /image/576/article/100x100 HTTP/1.1" 200 14776 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"
66.249.66.194 - - [22/Jan/2019:03:56:18 +0330] "GET /filter/b41,b665,c150%7C%D8%A8%D8%AE%D8%A7%D8%B1%D9%BE%D8%B2,p56 HTTP/1.1" 200 34277 "-" "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)" "-"
40.77.167.129 - - [22/Jan/2019:03:56:18 +0330] "GET /image/57710/productModel/100x100 HTTP/1.1" 200 1695 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"
207.46.13.136 - - [22/Jan/2019:03:56:18 +0330] "GET /product/10214 HTTP/1.1" 200 39677 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"
40.77.167.129 - - [22/Jan/2019:03:56:19 +0330] "GET /image/578/article/100x100 HTTP/1.1" 200 9831 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +http://www.bing.com/bingbot.htm)" "-"

Most websites on the Internet probably have a similar log. Many website admins and DevOps want to be able to process logs like this to do filtering and analytics.

After parsing with the framework there are 11 fields in the log:

7 string fields
4 integer fields

The whole list of fields and their data types is:

 "properties": {
  "remote_addr": {"type": "text"},
  "remote_user": {"type": "text"},
  "runtime": {"type": "long"},
  "time_local": {"type": "long"},
  "request_type": {"type": "text"},
  "request_path": {
    "type": "text",
    "fields": {
      "raw": {
        "type": "keyword"
      }
    }
  },
  "request_protocol": {"type": "text"},
  "status": {"type": "long"},
  "size": {"type": "long"},
  "referer": {"type": "text"},
  "usearagent": {"type": "text"}
}

We preliminarily convert the raw log to CSV so it’s easier to load to different databases and search engines.

Databases

So far we have made this test available for 3 databases:

Clickhouse - a powerful OLAP database,
Elasticsearch - general purpose “search and analytics engine”,
Manticore Search - “database for search”, Elasticsearch alternative.

In this test we make as little changes to database default settings as possible to not give either of them an unfair advantage. Testing at max tuning is no less important, but it's a subject for another benchmark. Here we want to understand what latency a regular non-experienced user can get after just installing a database and running it with its default settings. But to make it fair to compare one with another we still had to change a few settings:

Clickhouse: no tuning , just CREATE TABLE ... ENGINE = MergeTree() ORDER BY id and standard clickhouse-server docker image.
Elasticsearch: we test in 2 modes:
- with no tuning at all which is probably what most users do
- with number of shards equal to the number of CPU cores on the server, so Elasticsearch can utilize the CPUs more efficiently for lower response time, since as said in Elasticsearch official guide “Each shard runs the search on a single CPU thread”. The dataset size is only 3.5 GB, so it’s not clear if it’s required or not, but that’s why we are testing it.
- bootstrap.memory_lock=true since as said on https://www.elastic.co/guide/en/elasticsearch/reference/current/docker.html#_disable_swapping it needs to be done for performance.
- the docker image is standard
Manticore Search is used in a form of their official docker image + the columnar library they provide :
- we test Manticore’s default row-wise storage
- and columnar storage since Elasticsearch and Clickhouse don’t provide row-oriented stores and it may be more fair to compare with Manticore running in this mode.
- we added secondary_indexes = 1 to the config which enables secondary indexes while filtering (when loading data that’s built anyway). Since Elasticsearch uses secondary indexes by default and it’s fairly easy to enable the same in Manticore it makes sense to do it. Unfortunately in Clickhouse user would have to make quite an effort to do the same, hence it’s not done, since it would then be considered a heavy tuning which would then require further tuning of the other databases which would make things too complicated and unfair.
About caches

We've also configured the databases to not use any internal caches. Why this is important:
1. In this benchmark, we conduct an accurate latency measurement to find out what response time users can expect if they run one of the tested queries at a random moment, not after running the same query many times consequently.
2. Any cache is a shortcut to low latency. As written in Wikipedia "cache stores data so that future requests for that data can be served faster". But caches are different, they can be divided into 2 main groups:
  - 👌 those that just cache raw data stored on disk. For example many databases use mmap() to map the data stored on disk to memory, access it easily and let the operating system take care about the rest (reading it from disk when there's free memory, removing it from memory when it's needed for something more important etc). This is ok in terms of performance testing, because we let each database leverage the benefit of using the OS page cache (or its internal similar cache that just reads data from disk) That's exactly what we do in this benchmark.
  - ❗ those that are used to save results of previous calculations. And it's fine in many cases, but in terms of this benchmark letting database enable such a cache is a bad idea, because:
    - it breaks proper measuring: instead of measuring calculation time you start measuring how long it takes to find a value by a key in memory. It's not something we want to do in this test (but it's interesting in general and we'll perhaps do it in the future and publish some article "Benchmark of caches").
    - even if they save not a full result of a particular query, but results of its sub-calculations it's not good, because it breaks the idea of the test - "what response time users can expect if they run one of the tested queries at a random moment".
    - some databases have such a cache (it's usually called "query cache"), others don't so if we don't disable database internal caches we'll give an unfair advantage to those having that.
    So we do everything to make sure none of the database does this kind of caching.
What exactly we do to achieve that:

Clickhouse:
- SYSTEM DROP MARK CACHE, SYSTEM DROP UNCOMPRESSED CACHE, SYSTEM DROP COMPILED EXPRESSION CACHE before testing each new query (not each attempt of the same query).
Elasticsearch:
- "index.queries.cache.enabled": false in its configuration
- /_cache/clear?request=true&query=true&fielddata=true before testing each new query
  
  (not each attempt of the same query).
Manticore Search (in configuration file):
- qcache_max_bytes = 0
- docstore_cache_size = 0
Operating system:
- we do echo 3 > /proc/sys/vm/drop_caches; sync before
  
  each NEW query (NOT each attempt). I.e. for each new query we: - stop database - drop OS cache - start it back - make the very first cold query and measure its time - and make tens more attempts (up to 100 or until the coefficient of variation is low enough to consider the test results high quality)

Queries

The queries are mostly analytical that do filtering, sorting and grouping, but we’ve also included one full-text query which searches in the request URL:


[
"select avg(size) avg_size, status from logs10m group by status order by avg_size desc limit 20",
{
  "manticoresearch": "select count(*) as cnt, avg(runtime), avg(size) from logs10m where match('@request_path settings logo') order by cnt desc limit 20",
  "elasticsearch": "select count(*) as cnt, avg(runtime), avg(size) from logs10m where query('request_path settings logo') order by cnt desc limit 20",
  "clickhouse": "select count(*) as cnt, avg(runtime), avg(size) from logs10m where (match(request_path, '(?i)(\\W|\\A)settings\\Wlogo(\\W|\\z)') or match(request_path, '(?i)(\\W|\\A)logo\\Wsettings(\\W|\\z)')) limit 20 FORMAT JSON"
},
"select count(*) from logs10m",
"select count(*), avg(runtime) from logs10m group by status limit 20",
"select count(distinct request_path) cnt_distinct, status from logs10m group by status order by cnt_distinct desc limit 20",
"select min(size) min_size, status from logs10m group by status order by min_size desc, status desc limit 20",
"select request_path, runtime, status, size from logs10m where size > 0 order by runtime desc, size asc limit 20",
"select request_path, runtime, status, size, time_local from logs10m order by runtime desc, size desc, time_local desc limit 20",
"select status, count(*) from logs10m group by status order by count(*) desc limit 20",
"select status, sum(runtime) from logs10m group by status order by count(*) desc limit 20",
"select count(*) as cnt, request_path, avg(runtime), avg(size) from logs10m group by request_path order by cnt desc limit 20",
"select request_path, count(*), avg(runtime) runtime_avg, avg(size) from logs10m group by request_path order by runtime_avg desc limit 20"
]

Results

You can find all the results on the results page by selecting “Test: logs10m”.

Remember that the only high quality metric is “Fast avg” since it guarantees low coefficient of variation and high queries count conducted for each query. The other 2 (“Fastest” and “Slowest”) are provided with no guarantee since:

Slowest - is a single attempt result, in most cases the very first coldest query. Even though we purge OS cache before each cold query it can’t be considered stable. So it can be used for informational purposes only (even though many benchmark authors publish such results without any disclaimer).
Fastest - just the very fastest result, it should be in most cases similar to the “Fast avg” metric, but can be more volatile from run to run.

Remember the tests including the results are 100% transparent as well as everything in this project, so:

you can use the test framework to learn how they were made
and find raw test results in the results directory.

Unlike other less transparent and less objective benchmarks we are not making any conclusions, we are just leaving screenshots of the results here: