{"id":54918,"date":"2020-01-07T00:00:00","date_gmt":"2020-01-06T21:00:00","guid":{"rendered":"https:\/\/prohoster.info\/blog\/blog_prohoster\/ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb"},"modified":"2020-02-18T14:02:58","modified_gmt":"2020-02-18T11:02:58","slug":"ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb","status":"publish","type":"post","link":"https:\/\/prohoster.info\/cs\/blog\/administrirovanie\/ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb","title":{"rendered":"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"<p><noindex><a rel=\"nofollow\" href=\"https:\/\/clickhouse.yandex\/\">Clickhouse<\/a><\/noindex> \u2014 je sloupcov\u00fd syst\u00e9m spr\u00e1vy datab\u00e1z\u00ed pro online zpracov\u00e1n\u00ed analytick\u00fdch dotaz\u016f (OLAP) s otev\u0159en\u00fdm zdrojov\u00fdm k\u00f3dem, vytvo\u0159en\u00fd spole\u010dnost\u00ed Yandex. Pou\u017e\u00edv\u00e1 ho Yandex, CloudFlare, VK.com, Badoo a dal\u0161\u00ed slu\u017eby po cel\u00e9m sv\u011bt\u011b k ulo\u017een\u00ed skute\u010dn\u011b velk\u00fdch objem\u016f dat (vlo\u017een\u00edm tis\u00edc\u016f \u0159\u00e1dk\u016f za sekundu nebo petabajt\u016f dat ulo\u017een\u00fdch na disku).<\/p>\n<p>V b\u011b\u017en\u00edm, \"\u0159\u00e1dkov\u00e9m\" syst\u00e9mu spr\u00e1vy datab\u00e1z\u00ed, jeho\u017e p\u0159\u00edklady jsou MySQL, Postgres, MS SQL Server, jsou data uchov\u00e1v\u00e1na v tomto po\u0159ad\u00ed: <\/p>\n<p><img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/398f260de0152baf3cc11d884fa0d52d.png\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<br \/>\nP\u0159itom hodnoty, kter\u00e9 pat\u0159\u00ed k jednomu \u0159\u00e1dku, jsou fyzicky ulo\u017eeny vedle sebe. V sloupcov\u00fdch syst\u00e9mech spr\u00e1vy datab\u00e1z\u00ed jsou hodnoty z r\u016fzn\u00fdch sloupc\u016f ukl\u00e1d\u00e1ny odd\u011blen\u011b a data jednoho sloupce \u2013 dohromady:<\/p>\n<p><img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/938a2ee3d588ac866ec92bb3ea09c253.png\" style=\"display:block;margin: 0 auto;\" \/><noindex><a rel=\"nofollow\" name=\"habracut\"><\/a><\/noindex><\/p>\n<p>P\u0159\u00edklady sloupcov\u00fdch syst\u00e9m\u016f spr\u00e1vy datab\u00e1z\u00ed zahrnuj\u00ed Vertica, Paraccel (Actian Matrix, Amazon Redshift), Sybase IQ, Exasol, Infobright, InfiniDB, MonetDB (VectorWise, Actian Vector), LucidDB, SAP HANA, Google Dremel, Google PowerDrill, Druid, kdb+.<\/p>\n<p>Firma \u2013 mailforwarder <noindex><a rel=\"nofollow\" href=\"https:\/\/qwintry.com\/\">Qwintry<\/a><\/noindex> za\u010dala pou\u017e\u00edvat Clickhouse v roce 2018 pro sestavov\u00e1n\u00ed zpr\u00e1v a byla velmi ohromena jej\u00ed jednoduchost\u00ed, \u0161k\u00e1lovatelnost\u00ed, podporou SQL a rychlost\u00ed. Rychlost t\u00e9to datab\u00e1ze hrani\u010dila s kouzlem.<\/p>\n<h3>Jednoduchost<\/h3>\n<p>\nClickhouse se instaluje na Ubuntu jedin\u00fdm p\u0159\u00edkazem. Pokud zn\u00e1te SQL, m\u016f\u017eete okam\u017eit\u011b za\u010d\u00edt pou\u017e\u00edvat Clickhouse pro sv\u00e9 pot\u0159eby. To v\u0161ak neznamen\u00e1, \u017ee m\u016f\u017eete prov\u00e9st \u201eshow create table\u201c v MySQL a zkop\u00edrovat SQL do Clickhouse. <\/p>\n<p>Ve srovn\u00e1n\u00ed s MySQL existuj\u00ed v t\u00e9to datab\u00e1zi d\u016fle\u017eit\u00e9 rozd\u00edly v typech dat v definic\u00edch sch\u00e9mat tabulek, a proto v\u00e1m bude chv\u00edli trvat, ne\u017e uprav\u00edte definice sch\u00e9mat tabulek a prozkoum\u00e1te tabulkov\u00e9 enginy.<\/p>\n<p>Clickhouse skv\u011ble funguje bez jak\u00e9hokoli dal\u0161\u00edho softwaru, ale pokud chcete pou\u017e\u00edt replikaci, budete muset nainstalovat ZooKeeper. Anal\u00fdza v\u00fdkonu dotaz\u016f vykazuje vynikaj\u00edc\u00ed v\u00fdsledky - syst\u00e9mov\u00e9 tabulky obsahuj\u00ed v\u0161echny informace a v\u0161echna data lze z\u00edskat pomoc\u00ed star\u00e9ho a nudn\u00e9ho SQL.<\/p>\n<h3>V\u00fdkon<\/h3>\n<ul>\n<li><noindex><a rel=\"nofollow\" href=\"https:\/\/clickhouse.yandex\/benchmark.html#%5B%2522100000000%2522,%5B%2522ClickHouse%2522,%2522Vertica%2522,%2522MySQL%2522%5D,%5B%25220%2522,%25221%2522%5D%5D\">Benchmark<\/a><\/noindex> srovn\u00e1n\u00ed Clickhouse s Vertica a MySQL na serveru s konfigurac\u00ed: dva sockety Intel\u00ae Xeon\u00ae CPU E5-2650 v2 @ 2.60GHz; 128 GiB RAM; md RAID-5 na 8 6TB SATA HDD, ext4.<\/li>\n<li><noindex><a rel=\"nofollow\" href=\"https:\/\/www.altinity.com\/blog\/2017\/6\/20\/clickhouse-vs-redshift\">Benchmark<\/a><\/noindex> srovn\u00e1n\u00ed Clickhouse s cloudov\u00fdm datov\u00fdm \u00falo\u017ei\u0161t\u011bm Amazon RedShift.<\/li>\n<li>\u00daryvky z blogu <noindex><a rel=\"nofollow\" href=\"https:\/\/blog.cloudflare.com\/how-cloudflare-analyzes-1m-dns-queries-per-second\/\">Cloudflare o v\u00fdkonu Clickhouse<\/a><\/noindex>:<\/li>\n<\/ul>\n<p>\n<img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/b6ef9fbad8e8c864a227bcffd97ec0fe.png\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<br \/>\nDatab\u00e1ze ClickHouse m\u00e1 velmi jednoduch\u00fd design \u2014 v\u0161echny uzly v klastru maj\u00ed stejnou funk\u010dnost a k koordinaci pou\u017e\u00edvaj\u00ed pouze ZooKeeper. Vytvo\u0159ili jsme mal\u00fd klastr z n\u011bkolika uzl\u016f a provedli testov\u00e1n\u00ed, b\u011bhem kter\u00e9ho jsme zjistili, \u017ee syst\u00e9m m\u00e1 docela p\u016fsobiv\u00fd v\u00fdkon, kter\u00fd odpov\u00edd\u00e1 deklarovan\u00fdm v\u00fdhod\u00e1m v benchmarc\u00edch analytick\u00fdch datab\u00e1z\u00ed. Rozhodli jsme se podrobn\u011bji prozkoumat koncept, na kter\u00e9m ClickHouse stoj\u00ed. Prvn\u00edm p\u0159ek\u00e1\u017ekou pro v\u00fdzkum bylo nedostatek n\u00e1stroj\u016f a mal\u00fd po\u010det \u010dlen\u016f komunity ClickHouse, proto jsme se pono\u0159ili do designu t\u00e9to datab\u00e1ze, abychom pochopili, jak funguje.<\/p>\n<p>ClickHouse nepodporuje p\u0159\u00edjem dat p\u0159\u00edmo od Kafka, proto\u017ee jde pouze o datab\u00e1zi, tak\u017ee jsme napsali vlastn\u00ed slu\u017ebu adapt\u00e9r\u016f v jazyce Go. Ta \u010dte zak\u00f3dovan\u00e9 zpr\u00e1vy Cap\u2019n Proto z Kafka, p\u0159ev\u00e1d\u00ed je na TSV a vkl\u00e1d\u00e1 do ClickHouse po d\u00e1vk\u00e1ch p\u0159es HTTP rozhran\u00ed. Pozd\u011bji jsme tuto slu\u017ebu p\u0159epsali, abychom vyu\u017eili knihovnu Go spolu s vlastn\u00edm rozhran\u00edm ClickHouse pro zv\u00fd\u0161en\u00ed v\u00fdkonu. P\u0159i hodnocen\u00ed v\u00fdkonu p\u0159\u00edjmu d\u00e1vek jsme objevili d\u016fle\u017eitou v\u011bc \u2013 uk\u00e1zalo se, \u017ee v\u00fdkon ClickHouse siln\u011b z\u00e1vis\u00ed na velikosti d\u00e1vky, tedy po\u010dtu sou\u010dasn\u011b vkl\u00e1dan\u00fdch \u0159\u00e1dk\u016f. Abychom pochopili, pro\u010d se to d\u011bje, prozkoumali jsme, jak ClickHouse ukl\u00e1d\u00e1 data.<\/p>\n<p>Hlavn\u00edm enginem, sp\u00ed\u0161e rodinou engin\u016f pro tabulky, kter\u00e9 ClickHouse pou\u017e\u00edv\u00e1 k ukl\u00e1d\u00e1n\u00ed dat, je MergeTree. Tento engine je koncep\u010dn\u011b podobn\u00fd algoritmu LSM pou\u017e\u00edvan\u00e9mu v Google BigTable nebo Apache Cassandra, av\u0161ak vyh\u00fdb\u00e1 se vytv\u00e1\u0159en\u00ed mezi\u010dasu v pam\u011bti a zapisuje data p\u0159\u00edmo na disk. To mu poskytuje vynikaj\u00edc\u00ed propustnost z\u00e1pisu, proto\u017ee ka\u017ed\u00e1 vlo\u017een\u00e1 d\u00e1vka je t\u0159\u00edd\u011bna pouze podle \u201eprim\u00e1rn\u00edho kl\u00ed\u010de\u201c primary key, komprimov\u00e1na a zapisov\u00e1na na disk, aby vytvo\u0159ila segment. <\/p>\n<p>Absence of a memory table or any concept of data \"freshness\" also means that data can only be added; modification or deletion is not supported by the system. Currently, the only way to delete data is to remove it by calendar months, as segments never cross month boundaries. The ClickHouse team is actively working to make this feature customizable. On the other hand, this makes the writing and merging of segments conflict-free, so the throughput scales linearly with the number of parallel inserts until I\/O or core saturation occurs. <br \/>\nHowever, this circumstance also means that the system is not suitable for small batches, which is why Kafka services and inserters are used for buffering. Furthermore, ClickHouse continues to constantly merge segments in the background, so many small pieces of information will be combined and written multiple times, thus increasing write intensity. At the same time, too many unrelated pieces will trigger aggressive throttling of inserts until the merging is complete. We have found that the best compromise between real-time data ingestion and ingestion performance is to limit the number of inserts per second into the table.<\/p>\n<p>The key to reading performance of tables is indexing and data placement on disk. Regardless of how fast the processing is, when the engine has to scan terabytes of data from disk and only some of it is used, it will take time. ClickHouse is a columnar store, so each segment contains a file for each column with sorted values for each row. This way, entire columns that are not present in the query can be skipped initially, and then several cells can be processed in parallel with vectorized execution. To avoid full scans, each segment has a small index file. <\/p>\n<p>Vzhledem k tomu, \u017ee v\u0161echny sloupce jsou se\u0159azeny podle \u201eprim\u00e1rn\u00edho kl\u00ed\u010de\u201c, indexov\u00fd soubor obsahuje pouze zna\u010dky (zachycen\u00e9 \u0159\u00e1dky) ka\u017ed\u00e9ho N-t\u00e9ho \u0159\u00e1dku, aby je bylo mo\u017en\u00e9 uchov\u00e1vat v pam\u011bti i pro velmi velk\u00e9 tabulky. Nap\u0159\u00edklad mohou b\u00fdt v\u00fdchoz\u00ed nastaven\u00ed nastavena na \u201eozna\u010dit ka\u017ed\u00fd 8192. \u0159\u00e1dek\u201c, tak\u017ee \u201echud\u00e9\u201c indexov\u00e1n\u00ed tabulky s 1 trilionem \u0159\u00e1dk\u016f, kter\u00e9 se snadno vejde do pam\u011bti, zabere pouze 122 070 znak\u016f. <\/p>\n<h3>Rozvoj syst\u00e9mu<\/h3>\n<p>\nV\u00fdvoj a zdokonalov\u00e1n\u00ed Clickhouse lze sledovat na <noindex><a rel=\"nofollow\" href=\"https:\/\/github.com\/yandex\/ClickHouse\/pulse\">Github repozit\u00e1\u0159i<\/a><\/noindex> a ujistit se, \u017ee proces \u201ezr\u00e1n\u00ed\u201c prob\u00edh\u00e1 ohromuj\u00edc\u00edmi tempy. <\/p>\n<p><img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/19211da28e42178d26dff7623689df71.png\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<\/p>\n<h3>Popularita<\/h3>\n<p>\nZd\u00e1 se, \u017ee popularita Clickhouse roste exponenci\u00e1ln\u011b, zejm\u00e9na v rusky mluv\u00edc\u00ed komunit\u011b. Lo\u0148sk\u00e1 konference High Load 2018 (Moskva, 8-9. listopadu 2018) uk\u00e1zala, \u017ee takov\u00ed ob\u0159i jako vk.com a Badoo pou\u017e\u00edvaj\u00ed Clickhouse, prost\u0159ednictv\u00edm kter\u00e9ho vkl\u00e1daj\u00ed data (nap\u0159. logy) z des\u00edtek tis\u00edc server\u016f sou\u010dasn\u011b. Ve 40minutov\u00e9m videu <noindex><a rel=\"nofollow\" href=\"https:\/\/www.youtube.com\/watch?v=pbbcMcrQoXw\">Jury Nasretdinov z t\u00fdmu Vkontakte vysv\u011btluje, jak se to prov\u00e1d\u00ed<\/a><\/noindex>. Brzy zve\u0159ejn\u00edme p\u0159epis na Habr pro usnadn\u011bn\u00ed pr\u00e1ce s materi\u00e1lem.<\/p>\n<h3>Oblasti pou\u017eit\u00ed<\/h3>\n<p>\nPot\u00e9, co jsem str\u00e1vil n\u011bjak\u00fd \u010das studiem, mysl\u00edm, \u017ee existuj\u00ed oblasti, ve kter\u00fdch m\u016f\u017ee b\u00fdt ClickHouse u\u017eite\u010dn\u00fd nebo schopen zcela nahradit jin\u00e1, tradi\u010dn\u011bj\u0161\u00ed a popul\u00e1rn\u011bj\u0161\u00ed \u0159e\u0161en\u00ed, jako jsou MySQL, PostgreSQL, ELK, Google Big Query, Amazon RedShift, TimescaleDB, Hadoop, MapReduce, Pinot a Druid. D\u00e1le jsou uvedeny detaily pou\u017eit\u00ed ClickHouse pro modernizaci nebo \u00faplnou v\u00fdm\u011bnu v\u00fd\u0161e uveden\u00fdch DBMS. <\/p>\n<h3>Roz\u0161\u00ed\u0159en\u00ed mo\u017enost\u00ed MySQL a PostgreSQL<\/h3>\n<p>\nNed\u00e1vno jsme \u010d\u00e1ste\u010dn\u011b nahradili MySQL ClickHouse pro platformu informa\u010dn\u00edch bulletin\u016f <noindex><a rel=\"nofollow\" href=\"https:\/\/www.mautic.org\/\">Mautic newsletter<\/a><\/noindex>. Probl\u00e9m spo\u010d\u00edval v tom, \u017ee MySQL kv\u016fli nepromy\u0161len\u00e9mu designu zaznamen\u00e1val ka\u017edou odeslanou zpr\u00e1vu a ka\u017ed\u00fd odkaz v t\u00e9to zpr\u00e1v\u011b s hashov\u00e1n\u00edm base64, co\u017e vytvo\u0159ilo obrovskou tabulku MySQL (email_stats). Po odesl\u00e1n\u00ed pouze 10 milion\u016f e-mail\u016f na\u0161im p\u0159edplatitel\u016fm tato tabulka zauj\u00edmala 150 GB diskov\u00e9ho prostoru a MySQL za\u010dalo \u201ezpomalen\u011b\u201c reagovat na jednoduch\u00e9 dotazy. Abychom probl\u00e9m s diskov\u00fdm prostorem vy\u0159e\u0161ili, \u00fasp\u011b\u0161n\u011b jsme pou\u017eili kompresi tabulky InnoDB, kter\u00e1 ji zmen\u0161ila \u010dty\u0159ikr\u00e1t. Nicm\u00e9n\u011b nem\u00e1 smysl uchov\u00e1vat v\u00edce ne\u017e 20-30 milion\u016f e-mail\u016f v MySQL pouze pro \u010dten\u00ed historie, proto\u017ee ka\u017ed\u00fd jednoduch\u00fd dotaz, kter\u00fd z n\u011bjak\u00e9ho d\u016fvodu vy\u017eaduje \u00fapln\u00e9 skenov\u00e1n\u00ed, vede k swapov\u00e1n\u00ed a velk\u00e9mu zat\u00ed\u017een\u00ed I\/O, co\u017e je d\u016fvod, pro\u010d jsme pravideln\u011b dost\u00e1vali upozorn\u011bn\u00ed Zabbix.<\/p>\n<p><img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/11b3ee4b1a8b7c9457bd3871bd12049f.png\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<br \/>\nClickhouse pou\u017e\u00edv\u00e1 dva kompresn\u00ed algoritmy, kter\u00e9 zmen\u0161uj\u00ed objem dat p\u0159ibli\u017en\u011b o <noindex><a rel=\"nofollow\" href=\"https:\/\/www.altinity.com\/blog\/2017\/11\/21\/compression-in-clickhouse\">3-4 n\u00e1sobek<\/a><\/noindex>, ale v tomto konkr\u00e9tn\u00edm p\u0159\u00edpad\u011b byly data obzvl\u00e1\u0161\u0165 \u201ekomprimovateln\u00e9\u201c. <\/p>\n<p><img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/d192502073d57967620aa07da0bbfa0c.png\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<\/p>\n<h3>Nahrazen\u00ed ELK<\/h3>\n<p>\nNa z\u00e1klad\u011b vlastn\u00ed zku\u0161enosti vy\u017eaduje stack ELK (ElasticSearch, Logstash a Kibana, v tomto konkr\u00e9tn\u00edm p\u0159\u00edpad\u011b ElasticSearch) mnohem v\u00edce zdroj\u016f k provozu, ne\u017e je pot\u0159eba pro uchov\u00e1v\u00e1n\u00ed log\u016f. ElasticSearch je skv\u011bl\u00fd engine, pokud pot\u0159ebujete dobr\u00e9 full-textov\u00e9 vyhled\u00e1v\u00e1n\u00ed v logech (a j\u00e1 si nemysl\u00edm, \u017ee to opravdu pot\u0159ebujete), ale zaj\u00edmalo by m\u011b, pro\u010d se de facto stal standardn\u00edm engine pro veden\u00ed den\u00edku. Jeho v\u00fdkonnost p\u0159\u00edjmu v kombinaci s Logstash n\u00e1m vytv\u00e1\u0159ela probl\u00e9my i p\u0159i pom\u011brn\u011b mal\u00fdch z\u00e1t\u011b\u017e\u00edch a vy\u017eadovala st\u00e1le v\u011bt\u0161\u00ed objem opera\u010dn\u00ed pam\u011bti a diskov\u00e9ho prostoru. Jako datab\u00e1ze je Clickhouse lep\u0161\u00ed ne\u017e ElasticSearch z n\u00e1sleduj\u00edc\u00edch d\u016fvod\u016f:<\/p>\n<ul>\n<li>Podpora SQL dialektu;<\/li>\n<li>Lep\u0161\u00ed m\u00edra komprese ulo\u017een\u00fdch dat;<\/li>\n<li>Podpora hled\u00e1n\u00ed regul\u00e1rn\u00edch v\u00fdraz\u016f Regex m\u00edsto full-textov\u00e9ho hled\u00e1n\u00ed;<\/li>\n<li>Zlep\u0161en\u00e9 pl\u00e1nov\u00e1n\u00ed dotaz\u016f a vy\u0161\u0161\u00ed celkov\u00fd v\u00fdkon.<\/li>\n<\/ul>\n<p>\nV sou\u010dasn\u00e9 dob\u011b je nejv\u011bt\u0161\u00edm probl\u00e9mem p\u0159i srovn\u00e1n\u00ed ClickHouse s ELK nedostatek \u0159e\u0161en\u00ed pro export log\u016f a tak\u00e9 nedostatek dokumentace a v\u00fdukov\u00fdch materi\u00e1l\u016f k tomuto t\u00e9matu. Ka\u017ed\u00fd u\u017eivatel v\u0161ak m\u016f\u017ee nastavit ELK pomoc\u00ed p\u0159\u00edru\u010dky Digital Ocean, co\u017e je velmi d\u016fle\u017eit\u00e9 pro rychl\u00e9 zaveden\u00ed t\u011bchto technologi\u00ed. Existuje zde datab\u00e1zov\u00fd engine, ale zat\u00edm chyb\u00ed Filebeat pro ClickHouse. Ano, tam je p\u0159\u00edtomna <noindex><a rel=\"nofollow\" href=\"https:\/\/www.fluentd.org\/\">fluentd<\/a><\/noindex> a syst\u00e9m pro pr\u00e1ci s logy <noindex><a rel=\"nofollow\" href=\"https:\/\/github.com\/flant\/loghouse\">loghouse<\/a><\/noindex>, existuje n\u00e1stroj <noindex><a rel=\"nofollow\" href=\"https:\/\/github.com\/Altinity\/clicktail\">clicktail<\/a><\/noindex> pro import dat log\u016f do ClickHouse, ale to v\u0161e vy\u017eaduje v\u00edce \u010dasu. Nicm\u00e9n\u011b ClickHouse i tak vynik\u00e1 d\u00edky sv\u00e9 jednoduchosti, tak\u017ee i nov\u00e1\u010dci jej snadno nainstaluj\u00ed a p\u0159ejdou k plnohodnotn\u00e9mu pou\u017e\u00edv\u00e1n\u00ed doslova b\u011bhem 10 minut. <\/p>\n<p>Preferuj\u00edc minimalistick\u00e1 \u0159e\u0161en\u00ed, pokusil jsem se pou\u017e\u00edt FluentBit, n\u00e1stroj pro exportov\u00e1n\u00ed log\u016f s velmi malou pam\u011b\u0165ovou stopou, spolu s ClickHouse, p\u0159i\u010dem\u017e jsem se sna\u017eil vyhnout pou\u017eit\u00ed Kafka. Nicm\u00e9n\u011b je t\u0159eba odstranit mal\u00e9 neslu\u010ditelnosti, jako nap\u0159\u00edklad <noindex><a rel=\"nofollow\" href=\"https:\/\/github.com\/fluent\/fluent-bit\/issues\/848\">probl\u00e9my s form\u00e1tem data<\/a><\/noindex>, ne\u017e to bude mo\u017en\u00e9 ud\u011blat bez proxy vrstvy, kter\u00e1 p\u0159ev\u00e1d\u00ed data z FluentBit do ClickHouse.<\/p>\n<p>Jako alternativu k Kibana lze pou\u017e\u00edt ClickHouse jako backend <noindex><a rel=\"nofollow\" href=\"https:\/\/github.com\/Vertamedia\/clickhouse-grafana\">Grafana<\/a><\/noindex>. Jak jsem pochopil, p\u0159i tom mohou vznikat probl\u00e9my s v\u00fdkonem p\u0159i vykreslov\u00e1n\u00ed obrovsk\u00e9ho mno\u017estv\u00ed datov\u00fdch bod\u016f, zejm\u00e9na u star\u0161\u00edch verz\u00ed Grafana. V Qwintry jsme to zat\u00edm nezkou\u0161eli, ale st\u00ed\u017enosti na to se ob\u010das objevuj\u00ed na kan\u00e1lu podpory ClickHouse na Telegramu. <\/p>\n<h3>N\u00e1hrada za Google Big Query a Amazon RedShift (\u0159e\u0161en\u00ed pro velk\u00e9 firmy)<\/h3>\n<p>\nIde\u00e1ln\u00ed situace pro pou\u017eit\u00ed BigQuery je nahr\u00e1t 1 TB dat JSON a prov\u00e1d\u011bt nad nimi analytick\u00e9 dotazy. BigQuery je skv\u011bl\u00fd produkt, jeho\u017e \u0161k\u00e1lovatelnost je t\u011b\u017ek\u00e9 p\u0159ecenit. Je to mnohem slo\u017eit\u011bj\u0161\u00ed software ne\u017e ClickHouse, kter\u00fd b\u011b\u017e\u00ed na intern\u00edm clusteru, ale z pohledu klienta m\u00e1 mnoho spole\u010dn\u00e9ho s ClickHouse. BigQuery m\u016f\u017ee rychle \"zdra\u017eit\", jakmile za\u010dnete platit za ka\u017ed\u00fd SELECT, tak\u017ee je to skute\u010dn\u00e9 SaaS \u0159e\u0161en\u00ed se v\u0161emi sv\u00fdmi klady a z\u00e1pory. <\/p>\n<p>ClickHouse je nejlep\u0161\u00ed volba v p\u0159\u00edpad\u011b, \u017ee prov\u00e1d\u00edte mnoho n\u00e1ro\u010dn\u00fdch na v\u00fdpo\u010det dotaz\u016f. \u010c\u00edm v\u00edce SELECT dotaz\u016f provedete ka\u017ed\u00fd den \u2014 t\u00edm v\u00edce d\u00e1v\u00e1 smysl nahradit Big Query ClickHouse, proto\u017ee tato v\u00fdm\u011bna v\u00e1m m\u016f\u017ee u\u0161et\u0159it tis\u00edce dolar\u016f, pokud jde o mnoho terabajt\u016f zpracov\u00e1van\u00fdch dat. To neplat\u00ed pro ulo\u017een\u00e1 data, jejich\u017e zpracov\u00e1n\u00ed v Big Query je pom\u011brn\u011b levn\u00e9.<\/p>\n<p>V \u010dl\u00e1nku spoluzakladatele spole\u010dnosti Altinity Alexandra Zaitseva <noindex><a rel=\"nofollow\" href=\"https:\/\/www.altinity.com\/blog\/2017\/10\/23\/migration-to-clickhouse\">\u201eP\u0159echod na ClickHouse\u201c<\/a><\/noindex> se rozeb\u00edraj\u00ed v\u00fdhody takov\u00e9 migrace DBMS.<\/p>\n<h3>N\u00e1hrada za TimescaleDB<\/h3>\n<p>\nTimescaleDB je roz\u0161\u00ed\u0159en\u00edm PostgreSQL, kter\u00e9 optimalizuje pr\u00e1ci s \u010dasov\u00fdmi \u0159adami v b\u011b\u017en\u00e9 datab\u00e1zi.<noindex><a rel=\"nofollow\" href=\"https:\/\/docs.timescale.com\/v1.0\/introduction\">https:\/\/docs.timescale.com\/v1.0\/introduction<\/a><\/noindex>, <noindex><a rel=\"nofollow\" href=\"https:\/\/habr.com\/ru\/company\/zabbix\/blog\/458530\/\">https:\/\/habr.com\/ru\/company\/zabbix\/blog\/458530\/<\/a><\/noindex>). <\/p>\n<p>I kdy\u017e ClickHouse nen\u00ed v\u00e1\u017en\u00fdm konkurentem v oblasti \u010dasov\u00fdch \u0159ad, d\u00edky sloupcov\u00e9 struktu\u0159e a vektorov\u00e9mu vykon\u00e1v\u00e1n\u00ed dotaz\u016f je ve v\u011bt\u0161in\u011b p\u0159\u00edpad\u016f zpracov\u00e1n\u00ed analytick\u00fdch dotaz\u016f mnohem rychlej\u0161\u00ed ne\u017e TimescaleDB. V\u00fdkon p\u0159\u00edjmu d\u00e1vkov\u00fdch dat ClickHouse je p\u0159ibli\u017en\u011b t\u0159ikr\u00e1t vy\u0161\u0161\u00ed, nav\u00edc pou\u017e\u00edv\u00e1 dvacetkr\u00e1t m\u00e9n\u011b diskov\u00e9ho prostoru, co\u017e je opravdu d\u016fle\u017eit\u00e9 pro zpracov\u00e1n\u00ed velk\u00fdch objem\u016f historick\u00fdch dat.<noindex><a rel=\"nofollow\" href=\"https:\/\/www.altinity.com\/blog\/ClickHouse-for-time-series\">https:\/\/www.altinity.com\/blog\/ClickHouse-for-time-series<\/a><\/noindex>.<\/p>\n<p>Na rozd\u00edl od ClickHouse je jedin\u00fdm zp\u016fsobem, jak u\u0161et\u0159it trochu diskov\u00e9ho prostoru v TimescaleDB, pou\u017eit\u00ed ZFS nebo podobn\u00fdch souborov\u00fdch syst\u00e9m\u016f.<\/p>\n<p>Nadch\u00e1zej\u00edc\u00ed aktualizace ClickHouse pravd\u011bpodobn\u011b zavedou delta-kompresi, co\u017e jej u\u010din\u00ed je\u0161t\u011b l\u00e9pe p\u0159izp\u016fsoben\u00fdm pro zpracov\u00e1n\u00ed a uchov\u00e1v\u00e1n\u00ed dat \u010dasov\u00fdch \u0159ad. TimescaleDB m\u016f\u017ee b\u00fdt lep\u0161\u00ed volbou ne\u017e \"hol\u00fd\" ClickHouse v n\u00e1sleduj\u00edc\u00edch p\u0159\u00edpadech:<\/p>\n<ul>\n<li>mal\u00e9 instalace s velmi mal\u00fdm objemem RAM (&lt;3 GB);<\/li>\n<li>velk\u00e9 mno\u017estv\u00ed mal\u00fdch INSERT, kter\u00e9 nechcete bufferovat do velk\u00fdch fragment\u016f;<\/li>\n<li>lep\u0161\u00ed konzistence, jednotnost a po\u017eadavky ACID;<\/li>\n<li>podpora PostGIS;<\/li>\n<li>spojen\u00ed s existuj\u00edc\u00edmi tabulkami PostgreSQL, proto\u017ee v podstat\u011b Timescale DB je PostgreSQL.<\/li>\n<\/ul>\n<p><\/p>\n<h3>Konkurence se syst\u00e9my Hadoop a MapReduce<\/h3>\n<p>\nHadoop a dal\u0161\u00ed produkty MapReduce mohou vykon\u00e1vat mnoho slo\u017eit\u00fdch v\u00fdpo\u010dt\u016f, ale obvykle pracuj\u00ed s obrovsk\u00fdmi zpo\u017ed\u011bn\u00edmi. ClickHouse tento probl\u00e9m napravuje t\u00edm, \u017ee zpracov\u00e1v\u00e1 terabajty dat a t\u00e9m\u011b\u0159 okam\u017eit\u011b vrac\u00ed v\u00fdsledky. T\u00edm p\u00e1dem je ClickHouse mnohem efektivn\u011bj\u0161\u00ed pro rychl\u00e1 interaktivn\u00ed analytick\u00e1 vy\u0161et\u0159ov\u00e1n\u00ed, co\u017e by m\u011blo zaujmout specialisty na zpracov\u00e1n\u00ed dat.<\/p>\n<h3>Konkurence s Pinot a Druid<\/h3>\n<p>\nNejbli\u017e\u0161\u00edmi konkurenty ClickHouse jsou sloupcov\u00e9 line\u00e1rn\u011b \u0161k\u00e1lovateln\u00e9 open source produkty Pinot a Druid. Skv\u011bl\u00e1 pr\u00e1ce v porovn\u00e1n\u00ed t\u011bchto syst\u00e9m\u016f byla publikov\u00e1na v \u010dl\u00e1nku <noindex><a rel=\"nofollow\" href=\"https:\/\/medium.com\/@leventov\/comparison-of-the-open-source-olap-systems-for-big-data-ClickHouse-druid-and-pinot-8e042a5ed1c7\">Romana Leventa<\/a><\/noindex> z 1. \u00fanora 2018. <\/p>\n<p><img decoding=\"async\" alt=\"Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB\" src=\"\/wp-content\/uploads\/2020\/01\/fe58814aa97b287b70c2f724b72b448a.png\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<br \/>\nTento \u010dl\u00e1nek pot\u0159ebuje aktualizaci \u2013 tvrd\u00ed, \u017ee ClickHouse nepodporuje operace UPDATE a DELETE, co\u017e nen\u00ed zcela pravda o posledn\u00edch verz\u00edch.<\/p>\n<p>Nem\u00e1me dostatek zku\u0161enost\u00ed s t\u011bmito datab\u00e1zemi, ale v\u016fbec se mi nel\u00edb\u00ed slo\u017eitost zvolen\u00e9 infrastruktury, kter\u00e1 je pot\u0159ebn\u00e1 pro spu\u0161t\u011bn\u00ed Druid a Pinot \u2014 je to cel\u00fd bal\u00edk \"pohybliv\u00fdch \u010d\u00e1st\u00ed\" obklopen\u00fdch Javou ze v\u0161ech stran.<\/p>\n<p>Druid a Pinot jsou inkuba\u010dn\u00ed projekty Apache, jejich\u017e v\u00fdvoj je podrobn\u011b pops\u00e1n na str\u00e1nk\u00e1ch projekt\u016f GitHub. Pinot se objevil v inkub\u00e1toru v \u0159\u00edjnu 2018, zat\u00edmco Druid vznikl o 8 m\u011bs\u00edc\u016f d\u0159\u00edve \u2013 v \u00fanoru.<\/p>\n<p>Nedostatek informac\u00ed o tom, jak AFS funguje, ve mn\u011b vyvol\u00e1v\u00e1 ur\u010dit\u011b n\u011bkter\u00e9, mo\u017en\u00e1 hloup\u00e9 ot\u00e1zky. Zaj\u00edmalo by m\u011b, zda si auto\u0159i Pinot v\u0161imli, \u017ee Apache Foundation je v\u00edce naklon\u011bn\u00e1 Druid, a zda toto postaven\u00ed v\u016f\u010di konkurenci nevytvo\u0159ilo pocit z\u00e1visti? Zpomal\u00ed se v\u00fdvoj Druid a urychl\u00ed se v\u00fdvoj Pinot, pokud sponzo\u0159i podporuj\u00edc\u00ed prvn\u00ed, n\u00e1hle projev\u00ed z\u00e1jem o druh\u00fd?<\/p>\n<h3>Nev\u00fdhody ClickHouse<\/h3>\n<p>\nNedozr\u00e1lost: je z\u0159ejm\u00e9, \u017ee to je st\u00e1le nedosp\u011bl\u00e1 technologie, ale v ka\u017ed\u00e9m p\u0159\u00edpad\u011b to, co se zde objevuje, nen\u00ed patrn\u00e9 z jin\u00fdch sloupcov\u00fdch DBMS. <\/p>\n<p>Mal\u00e9 vkl\u00e1d\u00e1n\u00ed \u00fadaj\u016f nefunguje dob\u0159e p\u0159i vysok\u00e9 rychlosti: vkl\u00e1d\u00e1n\u00ed by m\u011blo b\u00fdt rozd\u011bleno na v\u011bt\u0161\u00ed kusy, proto\u017ee v\u00fdkon mal\u00fdch vkl\u00e1d\u00e1n\u00ed kles\u00e1 \u00fam\u011brn\u011b po\u010dtu sloupc\u016f v ka\u017ed\u00e9m \u0159\u00e1dku. Takto ClickHouse ukl\u00e1d\u00e1 data na disku \u2013 ka\u017ed\u00fd sloupec znamen\u00e1 1 nebo v\u00edce soubor\u016f, tak\u017ee k vlo\u017een\u00ed 1 \u0159\u00e1dku obsahuj\u00edc\u00edho 100 sloupc\u016f je nutn\u00e9 otev\u0159\u00edt a zapsat alespo\u0148 100 soubor\u016f. Proto je pro vyrovn\u00e1v\u00e1n\u00ed vkl\u00e1d\u00e1n\u00ed pot\u0159eba zprost\u0159edkovatel (pokud s\u00e1m klient nezaji\u0161\u0165uje vyrovn\u00e1v\u00e1n\u00ed) \u2013 obvykle je to Kafka nebo n\u011bjak\u00fd syst\u00e9m spr\u00e1vy front. Lze tak\u00e9 pou\u017e\u00edt engine Buffer table, abychom pozd\u011bji zkop\u00edrovali velk\u00e9 fragmenty dat do tabulek MergeTree.<\/p>\n<p>Spojen\u00ed tabulek jsou omezeny opera\u010dn\u00ed pam\u011bt\u00ed serveru, ale aspo\u0148 tam jsou! Nap\u0159\u00edklad Druid a Pinot v\u016fbec nemaj\u00ed takov\u00e1 spojen\u00ed, proto\u017ee je t\u011b\u017ek\u00e9 je implementovat v distribuovan\u00fdch syst\u00e9mech, kter\u00e9 nepodporuj\u00ed p\u0159enos velk\u00fdch blok\u016f dat mezi uzly. <\/p>\n<h3>Z\u00e1v\u011bry<\/h3>\n<p>\nV p\u0159\u00ed\u0161t\u00edch letech pl\u00e1nujeme \u0161irok\u00e9 vyu\u017eit\u00ed ClickHouse v Qwintry, proto\u017ee tato datab\u00e1zov\u00e1 SYST\u00c9MA poskytuje vynikaj\u00edc\u00ed rovnov\u00e1hu v\u00fdkonu, n\u00edzk\u00fdch n\u00e1klad\u016f, \u0161k\u00e1lovatelnosti a jednoduchosti. Jsem t\u00e9m\u011b\u0159 p\u0159esv\u011bd\u010den, \u017ee se rychle roz\u0161\u00ed\u0159\u00ed, jakmile komunit\u011b ClickHouse dojdou dal\u0161\u00ed zp\u016fsoby, jak ji vyu\u017e\u00edt v mal\u00fdch a st\u0159edn\u00edch instalac\u00edch.<\/p>\n<h3>Mal\u00e1 reklama \ud83d\ude42<\/h3>\n<p>\nD\u011bkujeme, \u017ee s n\u00e1mi z\u016fst\u00e1v\u00e1te. L\u00edb\u00ed se v\u00e1m na\u0161e \u010dl\u00e1nky? Chcete vid\u011bt v\u00edce zaj\u00edmav\u00fdch materi\u00e1l\u016f? Podpo\u0159te n\u00e1s objedn\u00e1vkou nebo doporu\u010den\u00edm zn\u00e1m\u00fdm, <noindex><a rel=\"nofollow\" href=\"https:\/\/ua-hosting.company\/cloudvps\/nl\">cloudov\u00e9 VPS pro v\u00fdvoj\u00e1\u0159e od 4,99 $<\/a><\/noindex>, <b>unik\u00e1tn\u00ed alternativa entry-level server\u016f, kterou jsme pro v\u00e1s vymysleli:<\/b> <noindex><a rel=\"nofollow\" href=\"https:\/\/habr.com\/company\/ua-hosting\/blog\/347386\/\">V\u0161echny pravdy o VPS (KVM) E5-2697 v3 (6 jader) 10GB DDR4 480GB SSD 1Gbps od 19 $ nebo jak spr\u00e1vn\u011b rozd\u011blit server?<\/a><\/noindex> (k dispozici varianty s RAID1 a RAID10, a\u017e 24 jader a a\u017e 40GB DDR4).<\/p>\n<p><b>Dell R730xd je dvakr\u00e1t levn\u011bj\u0161\u00ed v datov\u00e9m centru Equinix Tier IV v Amsterdamu?<\/b> Jen u n\u00e1s <b><noindex><a rel=\"nofollow\" href=\"https:\/\/ua-hosting.company\/serversnl\">2 x Intel TetraDeca-Core Xeon 2x E5-2697v3 2.6GHz 14C 64GB DDR4 4x960GB SSD 1Gbps 100 TB od 199 $<\/a><\/noindex> v Nizozemsku! <b>Dell R420 \u2014 2x E5-2430 2.2GHz 6C 128GB DDR3 2x960GB SSD 1Gbps 100TB \u2014 od 99 $!<\/b><\/b> \u010ct\u011bte o tom <noindex><a rel=\"nofollow\" href=\"https:\/\/habr.com\/company\/ua-hosting\/blog\/329618\/\">Jak postavit infrastrukturu podnikov\u00e9 t\u0159\u00eddy s pou\u017eit\u00edm server\u016f Dell R730xd E5-2650 v4 v hodnot\u011b 9000 eur za pakatel?<\/a><\/noindex><br \/>\n<br \/>Zdroj: <a content=\"nofollow\" rel=\"nofollow\" href=\"https:\/\/habr.com\/ru\/company\/ua-hosting\/blog\/483112\/\">habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"<p>Clickhouse \u2014 \u044d\u0442\u043e \u0441\u0442\u043e\u043b\u0431\u0446\u043e\u0432\u0430\u044f \u0441\u0438\u0441\u0442\u0435\u043c\u0430 \u0443\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u044f \u0431\u0430\u0437\u0430\u043c\u0438 \u0434\u0430\u043d\u043d\u044b\u0445 \u0434\u043b\u044f \u043e\u043d\u043b\u0430\u0439\u043d \u043e\u0431\u0440\u0430\u0431\u043e\u0442\u043a\u0438 \u0430\u043d\u0430\u043b\u0438\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u0437\u0430\u043f\u0440\u043e\u0441\u043e\u0432 (OLAP) \u0441 \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u043c \u0438\u0441\u0445\u043e\u0434\u043d\u044b\u043c \u043a\u043e\u0434\u043e\u043c, \u0441\u043e\u0437\u0434\u0430\u043d\u043d\u0430\u044f \u042f\u043d\u0434\u0435\u043a\u0441\u043e\u043c. \u0415\u0435 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u0443\u044e\u0442 \u042f\u043d\u0434\u0435\u043a\u0441, CloudFlare, VK.com, Badoo \u0438 \u0434\u0440\u0443\u0433\u0438\u0435 \u0441\u0435\u0440\u0432\u0438\u0441\u044b \u043f\u043e \u0432\u0441\u0435\u043c\u0443 \u043c\u0438\u0440\u0443 \u0434\u043b\u044f \u0445\u0440\u0430\u043d\u0435\u043d\u0438\u044f \u0434\u0435\u0439\u0441\u0442\u0432\u0438\u0442\u0435\u043b\u044c\u043d\u043e \u0431\u043e\u043b\u044c\u0448\u0438\u0445 \u043e\u0431\u044a\u0435\u043c\u043e\u0432 \u0434\u0430\u043d\u043d\u044b\u0445 (\u0432\u0441\u0442\u0430\u0432\u043a\u0430 \u0442\u044b\u0441\u044f\u0447 \u0441\u0442\u0440\u043e\u043a \u0432 \u0441\u0435\u043a\u0443\u043d\u0434\u0443 \u0438\u043b\u0438 \u043f\u0435\u0442\u0430\u0431\u0430\u0439\u0442\u044b \u0434\u0430\u043d\u043d\u044b\u0445, \u0445\u0440\u0430\u043d\u044f\u0449\u0438\u0445\u0441\u044f \u043d\u0430 \u0434\u0438\u0441\u043a\u0435). \u0412 \u043e\u0431\u044b\u0447\u043d\u043e\u0439, \u00ab\u0441\u0442\u0440\u043e\u043a\u043e\u0432\u043e\u0439\u00bb \u0421\u0423\u0411\u0414, \u043f\u0440\u0438\u043c\u0435\u0440\u0430\u043c\u0438 \u043a\u043e\u0442\u043e\u0440\u044b\u0445 [&hellip;]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[688],"tags":[],"class_list":["post-54918","post","type-post","status-publish","format-standard","hentry","category-administrirovanie"],"aioseo_notices":[],"aioseo_head":"\n\t\t<!-- All in One SEO 5.0.3 - aioseo.com -->\n\t<meta name=\"description\" content=\"Clickhouse \u2014 \u044d\u0442\u043e \u0441\u0442\u043e\u043b\u0431\u0446\u043e\u0432\u0430\u044f \u0441\u0438\u0441\u0442\u0435\u043c\u0430 \u0443\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u044f \u0431\u0430\u0437\u0430\u043c\u0438 \u0434\u0430\u043d\u043d\u044b\u0445 \u0434\u043b\u044f \u043e\u043d\u043b\u0430\u0439\u043d \u043e\u0431\u0440\u0430\u0431\u043e\u0442\u043a\u0438 \u0430\u043d\u0430\u043b\u0438\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u0437\u0430\u043f\u0440\u043e\u0441\u043e\u0432 (OLAP) \u0441 \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u043c \u0438\u0441\u0445\u043e\u0434\u043d\u044b\u043c \u043a\u043e\u0434\u043e\u043c.\" \/>\n\t<meta name=\"robots\" content=\"max-image-preview:large\" \/>\n\t<meta name=\"author\" content=\"Yuri Gagarin\"\/>\n\t<link rel=\"canonical\" href=\"https:\/\/prohoster.info\/cs\/blog\/administrirovanie\/ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb\" \/>\n\t\t<meta name=\"generator\" content=\"All in One SEO (AIOSEO) 5.0.3\" \/>\n\t\t<meta property=\"og:locale\" content=\"cs_CZ\" \/>\n\t\t<meta property=\"og:site_name\" content=\"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b\" \/>\n\t\t<meta property=\"og:type\" content=\"article\" \/>\n\t\t<meta property=\"og:title\" content=\"\ud83e\udd47\u0418\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u0435 Clickhouse \u0432 \u043a\u0430\u0447\u0435\u0441\u0442\u0432\u0435 \u0437\u0430\u043c\u0435\u043d\u044b ELK, Big Query \u0438 TimescaleDB | ProHoster\" \/>\n\t\t<meta property=\"og:description\" content=\"Clickhouse \u2014 \u044d\u0442\u043e \u0441\u0442\u043e\u043b\u0431\u0446\u043e\u0432\u0430\u044f \u0441\u0438\u0441\u0442\u0435\u043c\u0430 \u0443\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u044f \u0431\u0430\u0437\u0430\u043c\u0438 \u0434\u0430\u043d\u043d\u044b\u0445 \u0434\u043b\u044f \u043e\u043d\u043b\u0430\u0439\u043d \u043e\u0431\u0440\u0430\u0431\u043e\u0442\u043a\u0438 \u0430\u043d\u0430\u043b\u0438\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u0437\u0430\u043f\u0440\u043e\u0441\u043e\u0432 (OLAP) \u0441 \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u043c \u0438\u0441\u0445\u043e\u0434\u043d\u044b\u043c \u043a\u043e\u0434\u043e\u043c.\" \/>\n\t\t<meta property=\"og:url\" content=\"https:\/\/prohoster.info\/cs\/blog\/administrirovanie\/ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb\" \/>\n\t\t<meta property=\"og:image\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:secure_url\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:width\" content=\"350\" \/>\n\t\t<meta property=\"og:image:height\" content=\"350\" \/>\n\t\t<meta property=\"article:published_time\" content=\"2020-01-06T21:00:00+00:00\" \/>\n\t\t<meta property=\"article:modified_time\" content=\"2020-02-18T11:02:58+00:00\" \/>\n\t\t<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<meta property=\"article:author\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<!-- All in One SEO -->\n\n","aioseo_head_json":{"title":"\ud83e\udd47Pou\u017eit\u00ed Clickhouse jako n\u00e1hrady za ELK, Big Query a TimescaleDB | ProHoster","description":"Clickhouse je sloupcov\u00fd syst\u00e9m spr\u00e1vy datab\u00e1z\u00ed pro online zpracov\u00e1n\u00ed analytick\u00fdch dotaz\u016f (OLAP) s otev\u0159en\u00fdm zdrojov\u00fdm k\u00f3dem.","canonical_url":"https:\/\/prohoster.info\/cs\/blog\/administrirovanie\/ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"cs_CZ","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd47\u0418\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u0435 Clickhouse \u0432 \u043a\u0430\u0447\u0435\u0441\u0442\u0432\u0435 \u0437\u0430\u043c\u0435\u043d\u044b ELK, Big Query \u0438 TimescaleDB | ProHoster","og:description":"Clickhouse \u2014 \u044d\u0442\u043e \u0441\u0442\u043e\u043b\u0431\u0446\u043e\u0432\u0430\u044f \u0441\u0438\u0441\u0442\u0435\u043c\u0430 \u0443\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u044f \u0431\u0430\u0437\u0430\u043c\u0438 \u0434\u0430\u043d\u043d\u044b\u0445 \u0434\u043b\u044f \u043e\u043d\u043b\u0430\u0439\u043d \u043e\u0431\u0440\u0430\u0431\u043e\u0442\u043a\u0438 \u0430\u043d\u0430\u043b\u0438\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u0437\u0430\u043f\u0440\u043e\u0441\u043e\u0432 (OLAP) \u0441 \u043e\u0442\u043a\u0440\u044b\u0442\u044b\u043c \u0438\u0441\u0445\u043e\u0434\u043d\u044b\u043c \u043a\u043e\u0434\u043e\u043c.","og:url":"https:\/\/prohoster.info\/cs\/blog\/administrirovanie\/ispolzovanie-clickhouse-v-kachestve-zameny-elk-big-query-i-timescaledb","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2020-01-06T21:00:00+00:00","article:modified_time":"2020-02-18T11:02:58+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"54918","title":null,"description":null,"keywords":null,"keyphrases":null,"primary_term":null,"canonical_url":null,"og_title":null,"og_description":null,"og_object_type":"default","og_image_type":"default","og_image_url":null,"og_image_width":null,"og_image_height":null,"og_image_custom_url":null,"og_image_custom_fields":null,"og_video":null,"og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":false,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"","isEnabled":true},"graphs":[]},"schema_type":null,"schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":null,"robots_max_videopreview":null,"robots_max_imagepreview":"large","priority":null,"frequency":null,"local_seo":null,"seo_analyzer_scan_date":"2026-01-24 13:09:54","breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-02-28 19:55:36","updated":"2026-01-24 13:09:54","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/posts\/54918","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/comments?post=54918"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/posts\/54918\/revisions"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/media?parent=54918"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/categories?post=54918"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/cs\/wp-json\/wp\/v2\/tags?post=54918"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}