北京2021年(nian)10月(yue)21日(ri) /美(mei)通社/ -- 日(ri)前(qian)，亞馬遜云科技(ji)宣布Amazon Managed Service for Prometheus正式可(ke)(ke)用(yong)，這是一項可(ke)(ke)擴展、安全(quan)(quan)且高可(ke)(ke)用(yong)的服(fu)務，讓客(ke)(ke)戶可(ke)(ke)以(yi)更輕(qing)松地監控容器化應(ying)用(yong)程序。Amazon Managed Service for Prometheus與(yu)開(kai)源的Prometheus完全(quan)(quan)兼(jian)容，提供與(yu)客(ke)(ke)戶當前(qian)所(suo)使(shi)用(yong)的容器化應(ying)用(yong)監控工(gong)具同樣熟(shu)悉的時(shi)間(jian)序列數據模型(xing)和(he)(he)Prometheus Query Language（PromQL）。Amazon Managed Service for Prometheus是一項全(quan)(quan)面托(tuo)管的服(fu)務，會自動擴展所(suo)需的基礎設施，從容器化應(ying)用(yong)程序中提取、存儲和(he)(he)查詢運營(ying)指標。Amazon Managed Service for Prometheus與(yu)Amazon Identity and Access Management（IAM）和(he)(he)Amazon CloudTrail集(ji)成(cheng)，讓客(ke)(ke)戶更輕(qing)松地控制和(he)(he)審核對數據的訪問。使(shi)用(yong)Amazon Managed Service for Prometheus無(wu)需預付費(fei)(fei)用(yong)，客(ke)(ke)戶只需為提取、存儲和(he)(he)查詢的運營(ying)指標付費(fei)(fei)。欲開(kai)始(shi)使(shi)用(yong)Amazon Managed Service for Prometheus，請訪問。

容(rong)(rong)器因(yin)其(qi)能(neng)夠(gou)提高(gao)(gao)資(zi)(zi)源(yuan)(yuan)利用(yong)(yong)(yong)率并(bing)降低成(cheng)本，而受(shou)到越(yue)來(lai)越(yue)多(duo)客(ke)戶(hu)(hu)(hu)的(de)青睞，用(yong)(yong)(yong)于構(gou)建(jian)現代化(hua)(hua)應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序。容(rong)(rong)器通(tong)常運(yun)(yun)行(xing)(xing)(xing)時(shi)間(jian)短，共享(xiang)基(ji)礎(chu)設(she)(she)施(shi)資(zi)(zi)源(yuan)(yuan)，并(bing)產生大量(liang)的(de)運(yun)(yun)營指標(biao)，因(yin)此在(zai)使(shi)(shi)用(yong)(yong)(yong)傳(chuan)統面向裸機(ji)或(huo)虛(xu)擬服(fu)務器中運(yun)(yun)行(xing)(xing)(xing)的(de)應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序所設(she)(she)計的(de)使(shi)(shi)用(yong)(yong)(yong)工(gong)(gong)(gong)具(ju)時(shi)很難實現監控。客(ke)戶(hu)(hu)(hu)往往難以了(le)解大規模容(rong)(rong)器化(hua)(hua)應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序的(de)運(yun)(yun)行(xing)(xing)(xing)狀況和(he)(he)性能(neng)，當(dang)發(fa)生警(jing)報或(huo)出現問題時(shi)響應速度就(jiu)會很慢，從而導致應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序中斷，并(bing)損害(hai)用(yong)(yong)(yong)戶(hu)(hu)(hu)體驗。為了(le)克服(fu)這一(yi)挑戰(zhan)，許多(duo)客(ke)戶(hu)(hu)(hu)使(shi)(shi)用(yong)(yong)(yong)流行(xing)(xing)(xing)的(de)開(kai)源(yuan)(yuan)項目Prometheus 來(lai)監控其(qi)應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序。Prometheus經過優化(hua)(hua)，可(ke)存(cun)儲(chu)來(lai)自容(rong)(rong)器化(hua)(hua)應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序的(de)運(yun)(yun)營指標(biao)，客(ke)戶(hu)(hu)(hu)可(ke)輕松針對潛在(zai)應用(yong)(yong)(yong)程(cheng)(cheng)(cheng)序問題設(she)(she)置警(jing)報。為了(le)實現高(gao)(gao)可(ke)用(yong)(yong)(yong)性，在(zai)多(duo)臺(tai)服(fu)務器上安全地擴展 Prometheus并(bing)對其(qi)進(jin)(jin)行(xing)(xing)(xing)配置，需要(yao)大量(liang)手動(dong)工(gong)(gong)(gong)作，并(bing)且需要(yao)將其(qi)他(ta)開(kai)源(yuan)(yuan)工(gong)(gong)(gong)具(ju)和(he)(he)第三方(fang)工(gong)(gong)(gong)具(ju)拼接在(zai)一(yi)起。一(yi)旦Prometheus開(kai)始(shi)運(yun)(yun)行(xing)(xing)(xing)，客(ke)戶(hu)(hu)(hu)就(jiu)必(bi)須投入時(shi)間(jian)修補和(he)(he)升(sheng)級 Prometheus 以及(ji)他(ta)們日常使(shi)(shi)用(yong)(yong)(yong)的(de)其(qi)他(ta)工(gong)(gong)(gong)具(ju)。客(ke)戶(hu)(hu)(hu)還(huan)需要(yao)投入大量(liang)的(de)工(gong)(gong)(gong)程(cheng)(cheng)(cheng)時(shi)間(jian)優化(hua)(hua) Prometheus 如何(he)使(shi)(shi)用(yong)(yong)(yong)內存(cun)和(he)(he)存(cun)儲(chu)資(zi)(zi)源(yuan)(yuan)，縮短查詢響應時(shi)間(jian)并(bing)保持低成(cheng)本。這種復雜性和(he)(he)運(yun)(yun)營開(kai)銷(xiao)導致客(ke)戶(hu)(hu)(hu)需要(yao)進(jin)(jin)行(xing)(xing)(xing)權衡，一(yi)方(fang)面投入寶貴(gui)的(de)時(shi)間(jian)和(he)(he)資(zi)(zi)源(yuan)(yuan)維護其(qi)Prometheus部署(shu)及(ji)其(qi)支(zhi)持基(ji)礎(chu)設(she)(she)施(shi)，另一(yi)方(fang)面是代表(biao)最終(zhong)用(yong)(yong)(yong)戶(hu)(hu)(hu)進(jin)(jin)行(xing)(xing)(xing)構(gou)建(jian)和(he)(he)創(chuang)新(xin)。

Amazon Managed Service for Prometheus是一項適用于容(rong)器(qi)(qi)化(hua)應(ying)用程(cheng)(cheng)(cheng)序的(de)可(ke)擴展(zhan)、安全(quan)且高可(ke)用的(de)監控(kong)服務，可(ke)滿足應(ying)用程(cheng)(cheng)(cheng)序和(he)(he)基礎(chu)設施監控(kong)需求(qiu)自動管理提(ti)取、存儲和(he)(he)查(cha)詢(xun)運(yun)(yun)(yun)營指(zhi)標(biao)所需的(de)基礎(chu)設施。Amazon Managed Service for Prometheus與開源Prometheus完全(quan)兼容(rong)，客(ke)(ke)戶(hu)(hu)(hu)(hu)可(ke)以(yi)使(shi)用相同(tong)且熟悉的(de)時間序列數(shu)據模型和(he)(he)PromQL代碼來查(cha)詢(xun)運(yun)(yun)(yun)營指(zhi)標(biao)。客(ke)(ke)戶(hu)(hu)(hu)(hu)只需點擊幾(ji)下，即(ji)可(ke)從亞馬遜(xun)云(yun)科技容(rong)器(qi)(qi)服務（如Amazon Elastic Container Service、Amazon Elastic Kubernetes Service和(he)(he)Amazon Fargate）向(xiang)Amazon Managed Service for Prometheus發(fa)送運(yun)(yun)(yun)營指(zhi)標(biao)，享用Amazon Managed Service for Prometheus的(de)快速查(cha)詢(xun)響應(ying)和(he)(he)高性價比(bi)監控(kong)能力。通(tong)(tong)過使(shi)用Amazon Managed Service for Prometheus的(de)應(ying)用程(cheng)(cheng)(cheng)序編程(cheng)(cheng)(cheng)接口(kou)（API），客(ke)(ke)戶(hu)(hu)(hu)(hu)還可(ke)以(yi)更輕(qing)松、更安全(quan)地(di)從在云(yun)中或本地(di)運(yun)(yun)(yun)行(xing)(xing)的(de)現(xian)有(you)客(ke)(ke)戶(hu)(hu)(hu)(hu)管理的(de) Kubernetes 集群中提(ti)取運(yun)(yun)(yun)營指(zhi)標(biao)。如果客(ke)(ke)戶(hu)(hu)(hu)(hu)已有(you)Prometheus部署，或由于數(shu)據留存、延遲(chi)、監管或合(he)規性等需要在本地(di)運(yun)(yun)(yun)行(xing)(xing)Prometheus，他(ta)們可(ke)以(yi)通(tong)(tong)過單個命令行(xing)(xing)請求(qiu)輕(qing)松地(di)將其(qi) Prometheus 部署與Amazon Managed Service for Prometheus集成。Amazon Managed Service for Prometheus還內置IAM以(yi)控(kong)制訪問和(he)(he)權(quan)限，同(tong)時支(zhi)持(chi)Amazon CloudTrial審核用戶(hu)(hu)(hu)(hu)訪問以(yi)幫助客(ke)(ke)戶(hu)(hu)(hu)(hu)滿足企業安全(quan)與合(he)規要求(qiu)。Amazon Managed Service for Prometheus由亞馬遜(xun)云(yun)科技全(quan)面托(tuo)管，客(ke)(ke)戶(hu)(hu)(hu)(hu)無需構建、擴展(zhan)和(he)(he)維護 Prometheus 及(ji)其(qi)支(zhi)持(chi)基礎(chu)設施和(he)(he)工具，即(ji)可(ke)獲得針對容(rong)器(qi)(qi)化(hua)應(ying)用程(cheng)(cheng)(cheng)序優化(hua)的(de)高性價比(bi)監控(kong)服務。

亞馬遜云科技監控和可視化副總裁Nandini Ramani表示：“Prometheus因專(zhuan)為處理(li)容器化應用程序(xu)(xu)的需求而構建，而深(shen)受客(ke)戶(hu)青(qing)睞，但(dan)他們同(tong)時(shi)發(fa)現(xian)，自行(xing)(xing)大規模(mo)管理(li)和(he)運行(xing)(xing) Prometheus 既困(kun)難又耗時(shi)。借助Amazon Managed Service for Prometheus，客(ke)戶(hu)可以為其在(zai)亞馬遜云科技和(he)本地運行(xing)(xing)的容器化應用程序(xu)(xu)進(jin)行(xing)(xing)優化，獲(huo)得可擴(kuo)展(zhan)、安(an)全且(qie)高可用的監(jian)控服務。Amazon Managed Service for Prometheus消除了運行(xing)(xing) Prometheus 無差(cha)別的繁重工(gong)作，讓(rang)客(ke)戶(hu)專(zhuan)注于構建現(xian)代應用程序(xu)(xu)，幫(bang)助他們為最終用戶(hu)提供全新的創新體驗。”

Amazon Managed Service for Prometheus is available today in US East (Ohio), US East (N. Virginia), US West (Oregon), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Tokyo), Europe (Frankfurt), Europe (Ireland), and Europe (Stockholm), with availability in additional AWS Regions coming soon.

Amazon Managed Service for Prometheus目前已在美(mei)國東(dong)部(bu)（俄亥(hai)俄）、美(mei)國東(dong)部(bu)（弗吉尼亞(ya)北部(bu)）、美(mei)國西部(bu)（俄勒岡）、亞(ya)太(tai)(tai)地(di)區(qu)(qu)(qu)(qu)（新加坡）、亞(ya)太(tai)(tai)地(di)區(qu)(qu)(qu)(qu)（悉(xi)尼）、亞(ya)太(tai)(tai)地(di)區(qu)(qu)(qu)(qu)（東(dong)京）、歐(ou)洲（法蘭克福）、歐(ou)洲（愛爾(er)蘭）和歐(ou)洲（斯德哥爾(er)摩(mo)）區(qu)(qu)(qu)(qu)域(yu)正式可用，其他區(qu)(qu)(qu)(qu)域(yu)也將很(hen)快推(tui)出(chu)。

Fanatics是體育用品特許經銷商的全球領導者，它使用亞馬遜云科技的服務促進了如足球、橄欖球和籃球等各類體育運動球迷間的聯系。Fanatics平臺工程經理German Rodriguez表(biao)示：“我們(men)正重新評估我們(men)針對Amazon Elastic Kubernetes Service工作負載的(de)監控(kong)解決方案，并希(xi)望具(ju)有(you)(you)支(zhi)持(chi)開放標準的(de)可擴展性(xing)和高可用性(xing)。我們(men)之所以選擇Amazon Managed Service for Prometheus，因(yin)為它(ta)支(zhi)持(chi)開源(yuan)的(de)Prometheus，由亞馬(ma)(ma)遜云(yun)科技(ji)全面(mian)托管而具(ju)有(you)(you)可擴展性(xing)、安全且高可用。如今，我們(men)的(de)工程師可以更輕(qing)松(song)地(di)監控(kong)容器化應用程序的(de)性(xing)能，無論(lun)是部署在跨亞馬(ma)(ma)遜云(yun)科技(ji)還(huan)是在本地(di)，確保了為世界各地(di)的(de)球(qiu)迷提供最佳體驗。”

eTF1是法國最大廣播公司TF1旗下的OTT視頻服務。“我們最初自托管Prometheus，跨多個亞馬遜云科技賬戶，監控我們的Amazon Elastic Kubernetes Service工作負載，但很難維持一個高可用和高性能的監控環境。”eTF1基礎設施主管Ali Oubaziz表示，“Amazon Managed Service for Prometheus現(xian)在(zai)是我們主(zhu)要的(de)指(zhi)標監控(kong)平臺，使我們能(neng)夠大規模監控(kong)所有容(rong)器化工作負載 -- 即使在(zai)監控(kong)需求最(zui)高的(de)電(dian)視觀看高峰(feng)期也是如此。通過使用Amazon Managed Service for Prometheus，我們可以專(zhuan)注于提(ti)升用戶對OTT 電(dian)視服(fu)務的(de)滿意度。”