{"id":31230,"date":"2019-10-31T21:40:10","date_gmt":"2019-10-31T18:40:10","guid":{"rendered":"https:\/\/prohoster.info\/blog\/rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere\/"},"modified":"2019-10-31T21:40:10","modified_gmt":"2019-10-31T18:40:10","slug":"rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere","status":"publish","type":"post","link":"https:\/\/prohoster.info\/da\/blog\/administrirovanie\/rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere","title":{"rendered":"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"<p>I denne artikel vil jeg gerne tale om s\u00e6rtr\u00e6k ved arbejdet med All Flash-arrays fra AccelStor p\u00e5 en af de mest popul\u00e6re virtualiseringsplatforme \u2013 VMware vSphere. Is\u00e6r vil jeg fokusere p\u00e5 de parametre, der hj\u00e6lper med at opn\u00e5 maksimal effekt ved brug af et s\u00e5 kraftfuldt v\u00e6rkt\u00f8j som All Flash. <\/p>\n<p><\/p>\n<p><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/177f44f7e1c5f30c8c072af86b4f9999.jpg\" style=\"display:block;margin: 0 auto;\" \/><noindex><a rel=\"nofollow\" name=\"habracut\"><\/a><\/noindex><\/p>\n<p>All Flash-arrays fra AccelStor NeoSapphire\u2122 repr\u00e6senterer <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/page\/professional-models\">\u00e9n<\/a><\/noindex> eller <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/page\/modeli-vysokoj-dostupnosti-ha\">to<\/a><\/noindex> node-enheder baseret p\u00e5 SSD-lagringsenheder med en helt anden tilgang til implementering af datalagringskonceptet og organisering af adgangen til det ved hj\u00e6lp af egen teknologi <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/page\/flexiremap-technology\">FlexiRemap\u00ae<\/a><\/noindex> i stedet for de meget popul\u00e6re RAID-algoritmer. Arraysene giver blokadgang for v\u00e6rter gennem Fibre Channel- eller iSCSI-gr\u00e6nseflader. For retf\u00e6rdighedens skyld skal det n\u00e6vnes, at modeller med iSCSI-gr\u00e6nseflade ogs\u00e5 har filadgang som en god bonus. Men i denne artikel vil vi fokusere p\u00e5 brugen af blokprotokoller som de mest effektive for All Flash.<\/p>\n<p><\/p>\n<p>Hele processen med implementering og efterf\u00f8lgende konfiguration af samarbejdet mellem AccelStor-arrayet og virtualiseringssystemet VMware vSphere kan opdeles i flere faser:<\/p>\n<p><\/p>\n<ul>\n<li>Implementering af tilslutningstopologi og ops\u00e6tning af SAN-netv\u00e6rk;<\/li>\n<li>Konfiguration af All Flash-arrayet;<\/li>\n<li>Konfiguration af ESXi-v\u00e6rterne;<\/li>\n<li>Konfiguration af virtuelle maskiner.<\/li>\n<\/ul>\n<p><\/p>\n<p>Som udstyr til eksemplerne blev der anvendt AccelStor NeoSapphire\u2122-arrays med Fibre Channel- og iSCSI-gr\u00e6nseflader. Som grundl\u00e6ggende software \u2013 VMware vSphere 6.7U1.<\/p>\n<p><\/p>\n<p>F\u00f8r implementeringen af de systemer, der beskrives i artiklen, anbefales det kraftigt at gennemg\u00e5 dokumentationen fra VMware vedr\u00f8rende sp\u00f8rgsm\u00e5l om ydeevne (<noindex><a rel=\"nofollow\" href=\"https:\/\/www.vmware.com\/content\/dam\/digitalmarketing\/vmware\/en\/pdf\/techpaper\/performance\/vsphere-esxi-vcenter-server-67-performance-best-practices.pdf\">Performance Best Practices for VMware vSphere 6.7<\/a><\/noindex> ) og indstillinger for iSCSI (<noindex>Best Practices For Running VMware vSphere On iSCSI<\/noindex>)<\/p>\n<p><\/p>\n<h3><i><b>Tilslutningstopologi og ops\u00e6tning af SAN-netv\u00e6rk<\/b><\/i><\/h3>\n<p><\/p>\n<p>De vigtigste komponenter i SAN-netv\u00e6rket er HBA-adaptere i ESXi-v\u00e6rterne, SAN-switches og array-noder. En typisk topologi for et s\u00e5dant netv\u00e6rk vil se s\u00e5ledes ud:<\/p>\n<p><\/p>\n<p><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/0fa8f428d3120b551747e21d43d9095d.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/p>\n<p>Med termen Switch forst\u00e5s her b\u00e5de en separat fysisk switch eller et s\u00e6t af switches (Fabric) samt et delt enhed imellem forskellige tjenester (VSAN i tilf\u00e6lde af Fibre Channel og VLAN i tilf\u00e6lde af iSCSI). Brug af to uafh\u00e6ngige switches\/Fabric vil eliminere muligheden for et enkelt fejlpunk.<\/p>\n<p><\/p>\n<p>Direkte forbindelse af v\u00e6rter til arrayet underst\u00f8ttes, men det frar\u00e5des st\u00e6rkt. Ydeevnen fra All Flash arrays er meget h\u00f8j. For maksimal hastighed er det n\u00f8dvendigt at benytte alle porte i arrayet. Derfor er det obligatorisk at have mindst en switch mellem v\u00e6rterne og NeoSapphire\u2122.<\/p>\n<p><\/p>\n<p>At have to porte p\u00e5 HBA-v\u00e6rten er ogs\u00e5 et krav for at opn\u00e5 maksimal ydeevne og sikre fejltolerance.<\/p>\n<p><\/p>\n<p>Hvis Fibre Channel-gr\u00e6nsefladen anvendes, kr\u00e6ves der zoning-konfiguration for at undg\u00e5 mulige kollisioner mellem initiatorer og targets. Zonerne opbygges p\u00e5 grundlag af princippet \"\u00e9n initiatorport \u2013 en eller flere porte i arrayet\".<\/p>\n<p><\/p>\n<p>Hvis iSCSI-forbindelse anvendes, og en switch deles med andre tjenester, er det vigtigt at isolere iSCSI-trafikken inden for et separat VLAN. Det anbefales ogs\u00e5 st\u00e6rkt at aktivere Jumbo Frames (MTU = 9000) for at \u00f8ge st\u00f8rrelsen p\u00e5 pakkerne i netv\u00e6rket og dermed reducere m\u00e6ngden af overhead under transmission. Det er dog vigtigt at huske, at parameteren MTU skal \u00e6ndres p\u00e5 alle netv\u00e6rkets komponenter i k\u00e6den \"initiator-switch-target\" for korrekt funktion.<\/p>\n<p><\/p>\n<h3><i><b>Konfiguration af All Flash array<\/b><\/i><\/h3>\n<p><\/p>\n<p>Arrayet leveres til kunderne med allerede dannede grupper <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/page\/flexiremap-technology\">FlexiRemap\u00ae<\/a><\/noindex>. Derfor er det ikke n\u00f8dvendigt at foretage handlinger for at sammenflette de opbevaringsenheder i en enkelt struktur. Det er tilstr\u00e6kkeligt blot at oprette de n\u00f8dvendige volumener i den \u00f8nskede st\u00f8rrelse og m\u00e6ngde.<\/p>\n<p><\/p>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/zn\/x9\/nc\/znx9ncuyd2rgm8-jfvwf620zaqg.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/d09f9847203773280bc605be1c8fbc67.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><br \/>\n<noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/wx\/8l\/ua\/wx8lualewtdkixfwg0petpxbsre.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/ddcdaeb16263f79536db68f9c6ae2cff.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>For nemheds skyld er der en funktion til batchoprettelse af flere volumener med den specificerede st\u00f8rrelse p\u00e5 \u00e9n gang. Som standard oprettes \"tynde\" volumener, da dette g\u00f8r det muligt at udnytte den tilg\u00e6ngelige lagerplads mere effektivt (blandt andet takket v\u00e6re st\u00f8tte til Space Reclamation). Med hensyn til ydeevne er forskellen mellem \"tynde\" og \"tykke\" volumener ikke mere end 1%. Men hvis det er n\u00f8dvendigt at \"udnytte alle ressourcer\" fra arrayet, kan enhver \"tynd\" volumen altid konverteres til en \"tyk\". Men det skal huskes, at denne operation er irreversibel.<\/p>\n<p><\/p>\n<p>Hereafter, it is necessary to \"publish\" the created volumes and configure access rights for them on the part of the hosts using ACL (IP addresses for iSCSI and WWPN for FC) and physical separation by array ports. For iSCSI models, this is done by creating a Target.<\/p>\n<p><\/p>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/mw\/8y\/cg\/mw8ycgqngghoyuelortgysofswi.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/942d7ec437ad802ef828afa91bcde1f8.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><br \/>\n<noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/py\/xe\/28\/pyxe28gnrsxh_cx4ltjleogwsmw.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/f92d329b691d7b17380974253c8572ce.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>For FC models, publication occurs by creating a LUN for each port of the array.<\/p>\n<p><\/p>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/qh\/u3\/bx\/qhu3bxhm-kqinygguz298izbw7e.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/554aef268d6b18069edd3531b2429db5.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><br \/>\n<noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/oh\/al\/ey\/ohaleyllpb-0bmot9kawc4-ssyg.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/a770a4d3e6090c5bfaab3dad5487ee5a.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>To accelerate the configuration process, hosts can be grouped together. Moreover, if a multi-port FC HBA is used on the host (which is usually the case in practice), the system automatically determines that the ports of such an HBA belong to the same host due to WWPNs differing by one. Batch creation of Target\/LUN is also supported for both interfaces.<\/p>\n<p><\/p>\n<p>An important note when using the iSCSI interface is to create several targets for volumes at once to increase performance, as the queue on the target cannot be modified, and it will essentially be a bottleneck.<\/p>\n<p><\/p>\n<p><b><i><\/p>\n<h3>Configuring ESXi Hosts<\/h3>\n<p><\/i><\/b><\/p>\n<p><\/p>\n<p>From the perspective of ESXi hosts, the basic configuration is carried out according to a quite expected scenario. The sequence of actions for iSCSI connection:<\/p>\n<p><\/p>\n<ol>\n<li>Add Software iSCSI Adapter (not required if it's already added, or if using a Hardware iSCSI Adapter);<\/li>\n<li>Create a vSwitch through which the iSCSI traffic will pass and add physical uplink and VMkernel to it;<\/li>\n<li>Add the addresses of the array to Dynamic Discovery;<\/li>\n<li>Create a Datastore<\/li>\n<\/ol>\n<p><\/p>\n<p><b>Some important notes:<\/b><\/p>\n<p><\/p>\n<blockquote>\n<ul>\n<li>In general, it is certainly possible to use an existing vSwitch, but with a separate vSwitch, managing host settings will be significantly easier.<\/li>\n<li>It is necessary to separate Management traffic and iSCSI over separate physical links and\/or VLANs to avoid performance issues.<\/li>\n<li>The IP addresses of VMkernel and the corresponding ports of the All Flash array must be within the same subnet again due to performance issues.<\/li>\n<li>To ensure fault tolerance according to VMware rules, the vSwitch must have at least two physical uplinks.<\/li>\n<li>If Jumbo Frames are used, it is necessary to change the MTU for both the vSwitch and the VMkernel.<\/li>\n<li>Det vil ikke v\u00e6re overfl\u00f8digt at minde om, at if\u00f8lge VMware's anbefalinger for fysiske adaptere, der skal bruges til iSCSI trafik, er det n\u00f8dvendigt at konfigurere Teaming and Failover. Specifikt skal hver VMkernel kun arbejde gennem \u00e9n uplink, mens den anden uplink skal s\u00e6ttes til unused. For at sikre redundans skal to VMkernals tilf\u00f8jes, hver af dem skal arbejde gennem sin egen uplink.<\/li>\n<\/ul>\n<p>\n<\/p><\/blockquote>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/_h\/jb\/va\/_hjbvatnuhjwts4duy2epgb7bbi.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/b769fa87aa8552ed6dd775f9f9dcb1b5.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>VMkernel Adapter (vmk#)<br \/>\nFysisk Netv\u00e6rksadapter (vmnic#)<\/p>\n<p>vmk1 (Storage01)<br \/>\nAktive Adaptere<br \/>\nvmnic2<br \/>\nUbrugte Adaptere<br \/>\nvmnic3<\/p>\n<p>vmk2 (Storage02)<br \/>\nAktive Adaptere<br \/>\nvmnic3<br \/>\nUbrugte Adaptere<br \/>\nvmnic2<\/p>\n<p>Til forbindelse via Fibre Channel kr\u00e6ves der ingen forudg\u00e5ende handlinger. Du kan straks oprette et Datastore.<\/p>\n<p><\/p>\n<p>Efter oprettelsen af Datastore skal du sikre dig, at Round Robin-politikken anvendes for stierne til Target\/LUN som den mest effektive.<\/p>\n<p><\/p>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/zh\/zr\/t1\/zhzrt15baoc1qlp0w6ds-mys3hu.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/621474aa13123cd44518b263f222e6b5.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>Som standard planl\u00e6gger VMware at bruge denne politik efter m\u00f8nsteret: 1000 anmodninger via den f\u00f8rste sti, de n\u00e6ste 1000 anmodninger via den anden sti osv. S\u00e5dan interaktion mellem hosten og et to-controller array vil v\u00e6re ubalanceret. Derfor anbefaler vi at indstille Round Robin policy = 1 gennem Esxcli\/PowerCLI.<\/p>\n<p>\n<b class=\"spoiler_title\">Parametre<\/b><\/p>\n<p>For Esxcli:<\/p>\n<p><\/p>\n<ul>\n<li>Vis tilg\u00e6ngelige LUNs<\/li>\n<\/ul>\n<p><\/p>\n<p><b>esxcli storage nmp device list<\/b><\/p>\n<p><\/p>\n<ul>\n<li>Kopier enhedsnavn<\/li>\n<li>\u00c6ndre Round Robin Policy<\/li>\n<\/ul>\n<p><\/p>\n<p><b>esxcli storage nmp psp roundrobin deviceconfig set --type=iops --iops=1 --device=\u00abDevice_ID\u00bb<\/b><\/p>\n<p><\/p>\n<p>De fleste moderne applikationer er designet til at udveksle store datapakker for at maksimere b\u00e5ndbredden og reducere belastningen p\u00e5 CPU'en. Derfor sender ESXi som standard I\/O-anmodninger til storage-enheden i portioner op til 32767KB. Imidlertid vil det v\u00e6re mere effektivt i nogle scenarier at udveksle i mindre portioner. For AccelStor arrays er f\u00f8lgende scenarier relevante:<\/p>\n<p><\/p>\n<ul>\n<li>Den virtuelle maskine bruger UEFI i stedet for Legacy BIOS<\/li>\n<li>vSphere Replication anvendes<\/li>\n<\/ul>\n<p><\/p>\n<p>For s\u00e5danne scenarier anbefales det at \u00e6ndre parameteren Disk.DiskMaxIOSize til 4096.<\/p>\n<p><\/p>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/1g\/et\/3g\/1get3g8odkem1onrpkwjdilbnsi.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/f7072bb572580d9bd6077afab8c272ba.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>For iSCSI-forbindelser anbefales det at \u00e6ndre parameteren Login Timeout til 30 (som standard 5) for at forbedre forbindelsesstabiliteten og deaktivere forsinkelsen af bekr\u00e6ftelser af sendte pakker DelayedAck. Begge muligheder findes i vSphere Client: Host \u2192 Configure \u2192 Storage \u2192 Storage Adapters \u2192 Advanced Options for iSCSI adapteren.<\/p>\n<p><\/p>\n<p><noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/1y\/5_\/c8\/1y5_c80zsijgqtngo-9vt2zbbo8.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/e9700594d0074b5439dbb397c081dd48.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><br \/>\n<noindex><a rel=\"nofollow\" href=\"https:\/\/habrastorage.org\/webt\/bh\/yi\/tv\/bhyitvuujofnioaunwhzypx3fe0.png\"><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/c2dd6f3aaa6c0648763f66a94af2c251.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/a><\/noindex><\/p>\n<p><\/p>\n<p>Et v\u00e6sentligt punkt er antallet af anvendte volumes til datastore. Det er klart, at \u00f8nsket om at forenkle administrationen kan f\u00f8re til at oprette \u00e9n stor volume til hele kapaciteten af arrayet. Men det at have flere volumes, og dermed flere datastores, har en gavnlig effekt p\u00e5 den samlede ydeevne (mere om k\u00f8erne lidt l\u00e6ngere nede). Derfor anbefaler vi at oprette mindst to volumes.<\/p>\n<p><\/p>\n<p>For ikke s\u00e5 l\u00e6nge siden r\u00e5dgav VMware om at begr\u00e6nse antallet af virtuelle maskiner p\u00e5 \u00e9n datastore for igen at opn\u00e5 den h\u00f8jst mulige ydeevne. Men nu, is\u00e6r med udbredelsen af VDI, er dette problem ikke s\u00e5 presserende. Men det \u00e6ndrer ikke det gamle princip \u2013 at fordele virtuelle maskiner, der kr\u00e6ver intensiv I\/O, p\u00e5 forskellige datastores. Der er ikke noget bedre til at bestemme det optimale antal virtuelle maskiner p\u00e5 \u00e9n volume end at gennemf\u00f8re <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/page\/zaprosit-demo\">belastningstest af All Flash arrayet AccelStor <\/a><\/noindex>inden for din infrastruktur.<\/p>\n<p><\/p>\n<h3><i><b>Konfiguration af virtuelle maskiner<\/b><\/i><\/h3>\n<p><\/p>\n<p>Der er ingen s\u00e6rlige krav til ops\u00e6tning af virtuelle maskiner, eller rettere sagt de er ganske almindelige:<\/p>\n<p><\/p>\n<ul>\n<li>Brug den maksimalt mulige version af VM (kompatibilitet)<\/li>\n<li>V\u00e6r omhyggelig med at angive st\u00f8rrelsen af RAM ved t\u00e6t placering af virtuelle maskiner, for eksempel i VDI (da der som standard oprettes en sidefil, der er svarende med RAM-st\u00f8rrelsen ved opstart, hvilket bruger v\u00e6rdifuld kapacitet og p\u00e5virker den samlede ydeevne)<\/li>\n<li>Brug de mest ydeevnedygtige versioner af adapterne til I\/O: netv\u00e6rkstype VMXNET 3 og SCSI-type PVSCSI<\/li>\n<li>Brug typen disk Thick Provision Eager Zeroed for maksimal ydeevne og Thin Provisioning for maksimal effektiv udnyttelse af lagerplads<\/li>\n<li>Begr\u00e6ns om muligt driften af I\/O-uskadelige maskiner ved hj\u00e6lp af Virtual Disk Limit<\/li>\n<li>Installer altid VMware Tools <\/li>\n<\/ul>\n<p><\/p>\n<h3><i><b>Bem\u00e6rkninger om k\u00f8erne<\/b><\/i><\/h3>\n<p><\/p>\n<p>K\u00f8 (eller Outstanding I\/Os) er antallet af ind-\/uddataanmodninger (SCSI-kommandoer), der venter p\u00e5 behandling p\u00e5 et givet tidspunkt for et specifikt device\/applikation. I tilf\u00e6lde af, at k\u00f8en overbelastes, udstedes fejlmeddelelser om QFULL, hvilket i sidste ende resulterer i en stigning i latenstid. Ved brug af diskbaserede (spindel) lagringssystemer g\u00e6lder det teoretisk, at jo h\u00f8jere k\u00f8en er, desto h\u00f8jere er ydeevnen. Dog b\u00f8r man v\u00e6re varsom, da man let kan ramme QFULL. I tilf\u00e6lde af All Flash systemer er det p\u00e5 den ene side lidt lettere, da arrayet har latenser, der er ordentlig lavere, og derfor kr\u00e6ves det ofte ikke at justere st\u00f8rrelsen p\u00e5 k\u00f8erne separat. P\u00e5 den anden side kr\u00e6ver nogle brugs-scenarier (st\u00e6rk ubalance i IO-kravene for specifikke virtuelle maskiner, test af maksimal ydeevne osv.) om ikke at \u00e6ndre p\u00e5 k\u00f8parametrene, s\u00e5 i det mindste forst\u00e5, hvilke resultater der kan opn\u00e5s, og, vigtigst af alt, p\u00e5 hvilke m\u00e5der.<\/p>\n<p><\/p>\n<p>P\u00e5 selve All Flash arrayet AccelStor er der ingen gr\u00e6nser med hensyn til volumener eller ind-\/uddata porte. Om n\u00f8dvendigt kan selv et enkelt volum f\u00e5 alle ressourcerne fra arrayet. Den eneste begr\u00e6nsning p\u00e5 k\u00f8erne findes ved iSCSI targets. Derfor blev der tidligere n\u00e6vnt et behov for at oprette flere (ideelt op til 8 stk.) targets pr. volum for at overvinde denne gr\u00e6nse. Derudover skal det gentages, at AccelStor arrayene er yderst effektive l\u00f8sninger. Derfor b\u00f8r alle interface-porte i systemet anvendes for at opn\u00e5 maksimal hastighed.<\/p>\n<p><\/p>\n<p>Fra ESXi v\u00e6rtens perspektiv er situationen helt anderledes. Selve v\u00e6rten anvender en praksis, hvor ressourceadgang er ligelig for alle deltagere. Derfor findes der separate IO-k\u00f8er til g\u00e6ste-OS og HBA. K\u00f8erne til g\u00e6ste-OS kombineres fra k\u00f8erne til den virtuelle SCSI-adapter og den virtuelle disk:<\/p>\n<p><\/p>\n<p><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/e6033427e6f554490c9a6460574b7f6e.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/p>\n<p>K\u00f8en til HBA afh\u00e6nger af den specifikke type\/leverand\u00f8r:<\/p>\n<p><\/p>\n<p><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/a0b078e63bdf8f306a25bfecd1310cb0.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/p>\n<p>Den endelige ydeevne for den virtuelle maskine vil blive bestemt af den laveste v\u00e6rdi af k\u00f8dybdegr\u00e6nsen (Queue Depth limit) blandt hostkomponenterne. <\/p>\n<p><\/p>\n<p>Takket v\u00e6re disse v\u00e6rdier kan vi vurdere de ydeevneindikatorer, som vi kan opn\u00e5 i en given konfiguration. For eksempel \u00f8nsker vi at finde den teoretiske ydeevne for en virtuel maskine (uden henvisning til blokken) med en latency p\u00e5 0,5 ms. S\u00e5 vil dens IOPS = (1.000\/latency) * Outstanding I\/Os (Queue Depth limit)<\/p>\n<p>\n<b class=\"spoiler_title\">Eksempler<\/b><\/p>\n<p><b>Eksempel 1<\/b><\/p>\n<p><\/p>\n<ul>\n<li>FC Emulex HBA Adapter<\/li>\n<li>\u00c9n VM p\u00e5 datastore<\/li>\n<li>VMware Paravirtual SCSI Adapter<\/li>\n<\/ul>\n<p><\/p>\n<p>Her defineres Queue Depth limit af Emulex HBA. Derfor er IOPS = (1000\/0,5)*32 = 64K<\/p>\n<p><\/p>\n<p><b>Eksempel 2<\/b><\/p>\n<p><\/p>\n<ul>\n<li>VMware iSCSI Software Adapter<\/li>\n<li>\u00c9n VM p\u00e5 datastore<\/li>\n<li>VMware Paravirtual SCSI Adapter<\/li>\n<\/ul>\n<p><\/p>\n<p>Her defineres Queue Depth limit allerede af Paravirtual SCSI Adapter. Derfor er IOPS = (1000\/0,5)*64 = 128K<\/p>\n<p><\/p>\n<p>Topmodellerne af All Flash arrays fra AccelStor (for eksempel, <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/product\/neosapphire-p710\">P710<\/a><\/noindex>) kan levere 700K IOPS til skrivning ved 4K-blokke. Med en s\u00e5dan blokst\u00f8rrelse er det helt klart, at \u00e9n virtuel maskine ikke kan belaste en s\u00e5dan array. Dette kr\u00e6ver 11 (for eksempel 1) eller 6 (for eksempel 2) virtuelle maskiner.<\/p>\n<p><\/p>\n<p>Resultatet er, at med den rette ops\u00e6tning af alle de beskrevne komponenter i det virtuelle datacenter kan man opn\u00e5 ganske imponerende resultater med hensyn til ydeevne.<\/p>\n<p><\/p>\n<p><img decoding=\"async\" alt=\"Anbefalinger til konfiguration af AFA AccelStor ved brug af VMware vSphere\" src=\"\/wp-content\/uploads\/2019\/04\/eee93cccf16539410695c6f79d94deda.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/p>\n<p><i>4K Random, 70% L\u00e6s\/30% Skriv<\/i><\/p>\n<p><\/p>\n<p>I virkeligheden er verden meget mere kompleks, end man kan beskrive med en simpel formel. P\u00e5 \u00e9n v\u00e6rt er der altid mange virtuelle maskiner med forskellige konfigurationer og krav til IO. Og behandlingen af input\/output udf\u00f8res af v\u00e6rtens processor, hvis kapacitet ikke er uendelig. S\u00e5 for at udnytte det fulde potentiale af den samme <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/product\/neosapphire-p710\">model P710<\/a><\/noindex> i virkeligheden kr\u00e6ves der tre v\u00e6rter. Derudover bidrager applikationer, der k\u00f8rer inde i de virtuelle maskiner, med deres egne korrektioner. Derfor anbefaler vi for pr\u00e6cisions sizing <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/page\/zaprosit-demo\">at bruge testmodeller i praksis<\/a><\/noindex> All Flash arrays <noindex><a rel=\"nofollow\" href=\"http:\/\/accelstor.ru\/\">AccelStor<\/a><\/noindex> inde i kundens infrastruktur med reelle nuv\u00e6rende opgaver.<\/p>\n<p>Kilde: <a content=\"nofollow\" rel=\"nofollow\" href=\"https:\/\/habr.com\/ru\/company\/accelstor\/blog\/447390\/\">habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"<p>\u0412 \u0440\u0430\u043c\u043a\u0430\u0445 \u0434\u0430\u043d\u043d\u043e\u0439 \u0441\u0442\u0430\u0442\u044c\u0438 \u0445\u043e\u0442\u0435\u043b\u043e\u0441\u044c \u0431\u044b \u0440\u0430\u0441\u0441\u043a\u0430\u0437\u0430\u0442\u044c \u043e\u0431 \u043e\u0441\u043e\u0431\u0435\u043d\u043d\u043e\u0441\u0442\u044f\u0445 \u0440\u0430\u0431\u043e\u0442\u044b All Flash \u043c\u0430\u0441\u0441\u0438\u0432\u043e\u0432 AccelStor \u0441 \u043e\u0434\u043d\u043e\u0439 \u0438\u0437 \u043f\u043e\u043f\u0443\u043b\u044f\u0440\u043d\u0435\u0439\u0448\u0438\u0445 \u043f\u043b\u0430\u0442\u0444\u043e\u0440\u043c \u0432\u0438\u0440\u0442\u0443\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u2013 VMware vSphere. \u0412 \u0447\u0430\u0441\u0442\u043d\u043e\u0441\u0442\u0438, \u0430\u043a\u0446\u0435\u043d\u0442\u0438\u0440\u043e\u0432\u0430\u0442\u044c \u0432\u043d\u0438\u043c\u0430\u043d\u0438\u0435 \u043d\u0430 \u0442\u0435\u0445 \u043f\u0430\u0440\u0430\u043c\u0435\u0442\u0440\u0430\u0445, \u043a\u043e\u0442\u043e\u0440\u044b\u0435 \u043f\u043e\u043c\u043e\u0433\u0443\u0442 \u043f\u043e\u043b\u0443\u0447\u0438\u0442\u044c \u043c\u0430\u043a\u0441\u0438\u043c\u0430\u043b\u044c\u043d\u044b\u0439 \u044d\u0444\u0444\u0435\u043a\u0442 \u043e\u0442 \u0438\u0441\u043f\u043e\u043b\u044c\u0437\u043e\u0432\u0430\u043d\u0438\u044f \u0442\u0430\u043a\u043e\u0433\u043e \u043c\u043e\u0449\u043d\u043e\u0433\u043e \u0438\u043d\u0441\u0442\u0440\u0443\u043c\u0435\u043d\u0442\u0430, \u043a\u0430\u043a All Flash. All Flash \u043c\u0430\u0441\u0441\u0438\u0432\u044b AccelStor NeoSapphire\u2122 \u043f\u0440\u0435\u0434\u0441\u0442\u0430\u0432\u043b\u044f\u044e\u0442 \u0441\u043e\u0431\u043e\u0439 \u043e\u0434\u043d\u043e \u0438\u043b\u0438 \u0434\u0432\u0443\u0445 \u043d\u043e\u0434\u043e\u0432\u044b\u0435 \u0443\u0441\u0442\u0440\u043e\u0439\u0441\u0442\u0432\u0430 [&hellip;]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":23196,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[688],"tags":[],"class_list":["post-31230","post","type-post","status-publish","format-standard","has-post-thumbnail","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=\"\u0412 \u0440\u0430\u043c\u043a\u0430\u0445 \u0434\u0430\u043d\u043d\u043e\u0439 \u0441\u0442\u0430\u0442\u044c\u0438 \u0445\u043e\u0442\u0435\u043b\u043e\u0441\u044c \u0431\u044b \u0440\u0430\u0441\u0441\u043a\u0430\u0437\u0430\u0442\u044c \u043e\u0431 \u043e\u0441\u043e\u0431\u0435\u043d\u043d\u043e\u0441\u0442\u044f\u0445 \u0440\u0430\u0431\u043e\u0442\u044b All Flash \u043c\u0430\u0441\u0441\u0438\u0432\u043e\u0432 AccelStor \u0441 \u043e\u0434\u043d\u043e\u0439 \u0438\u0437 \u043f\u043e\u043f\u0443\u043b\u044f\u0440\u043d\u0435\u0439\u0448\u0438\u0445 \u043f\u043b\u0430\u0442\u0444\u043e\u0440\u043c \u0432\u0438\u0440\u0442\u0443\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u2013 VMware vSphere.\" \/>\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\/da\/blog\/administrirovanie\/rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere\" \/>\n\t\t<meta name=\"generator\" content=\"All in One SEO (AIOSEO) 5.0.3\" \/>\n\t\t<meta property=\"og:locale\" content=\"da_DK\" \/>\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\u0420\u0435\u043a\u043e\u043c\u0435\u043d\u0434\u0430\u0446\u0438\u0438 \u043f\u043e \u043d\u0430\u0441\u0442\u0440\u043e\u0439\u043a\u0435 AFA AccelStor \u043f\u0440\u0438 \u0440\u0430\u0431\u043e\u0442\u0435 \u0441 VMware vSphere | ProHoster\" \/>\n\t\t<meta property=\"og:description\" content=\"\u0412 \u0440\u0430\u043c\u043a\u0430\u0445 \u0434\u0430\u043d\u043d\u043e\u0439 \u0441\u0442\u0430\u0442\u044c\u0438 \u0445\u043e\u0442\u0435\u043b\u043e\u0441\u044c \u0431\u044b \u0440\u0430\u0441\u0441\u043a\u0430\u0437\u0430\u0442\u044c \u043e\u0431 \u043e\u0441\u043e\u0431\u0435\u043d\u043d\u043e\u0441\u0442\u044f\u0445 \u0440\u0430\u0431\u043e\u0442\u044b All Flash \u043c\u0430\u0441\u0441\u0438\u0432\u043e\u0432 AccelStor \u0441 \u043e\u0434\u043d\u043e\u0439 \u0438\u0437 \u043f\u043e\u043f\u0443\u043b\u044f\u0440\u043d\u0435\u0439\u0448\u0438\u0445 \u043f\u043b\u0430\u0442\u0444\u043e\u0440\u043c \u0432\u0438\u0440\u0442\u0443\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u2013 VMware vSphere.\" \/>\n\t\t<meta property=\"og:url\" content=\"https:\/\/prohoster.info\/da\/blog\/administrirovanie\/rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere\" \/>\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=\"2019-10-31T18:40:10+00:00\" \/>\n\t\t<meta property=\"article:modified_time\" content=\"2019-10-31T18:40:10+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\udd47Anbefalinger til ops\u00e6tning af AFA AccelStor ved arbejde med VMware vSphere | ProHoster","description":"I denne artikel vil jeg gerne fort\u00e6lle om de unikke egenskaber ved All Flash arrays fra AccelStor med en af de mest popul\u00e6re virtualiseringsplatforme \u2013 VMware vSphere.","canonical_url":"https:\/\/prohoster.info\/da\/blog\/administrirovanie\/rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"da_DK","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\u0420\u0435\u043a\u043e\u043c\u0435\u043d\u0434\u0430\u0446\u0438\u0438 \u043f\u043e \u043d\u0430\u0441\u0442\u0440\u043e\u0439\u043a\u0435 AFA AccelStor \u043f\u0440\u0438 \u0440\u0430\u0431\u043e\u0442\u0435 \u0441 VMware vSphere | ProHoster","og:description":"\u0412 \u0440\u0430\u043c\u043a\u0430\u0445 \u0434\u0430\u043d\u043d\u043e\u0439 \u0441\u0442\u0430\u0442\u044c\u0438 \u0445\u043e\u0442\u0435\u043b\u043e\u0441\u044c \u0431\u044b \u0440\u0430\u0441\u0441\u043a\u0430\u0437\u0430\u0442\u044c \u043e\u0431 \u043e\u0441\u043e\u0431\u0435\u043d\u043d\u043e\u0441\u0442\u044f\u0445 \u0440\u0430\u0431\u043e\u0442\u044b All Flash \u043c\u0430\u0441\u0441\u0438\u0432\u043e\u0432 AccelStor \u0441 \u043e\u0434\u043d\u043e\u0439 \u0438\u0437 \u043f\u043e\u043f\u0443\u043b\u044f\u0440\u043d\u0435\u0439\u0448\u0438\u0445 \u043f\u043b\u0430\u0442\u0444\u043e\u0440\u043c \u0432\u0438\u0440\u0442\u0443\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u2013 VMware vSphere.","og:url":"https:\/\/prohoster.info\/da\/blog\/administrirovanie\/rekomendatsii-po-nastrojke-afa-accelstor-pri-rabote-s-vmware-vsphere","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":"2019-10-31T18:40:10+00:00","article:modified_time":"2019-10-31T18:40:10+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"31230","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":"2026-01-21 05:09:19","breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-03-01 03:20:31","updated":"2026-01-21 05:09:19","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/posts\/31230","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/comments?post=31230"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/posts\/31230\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/media\/23196"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/media?parent=31230"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/categories?post=31230"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/da\/wp-json\/wp\/v2\/tags?post=31230"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}