[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC):2025-08-17。"],[],[],null,["# Optimize performance\n\nThis page provides details on how you can optimize Google Cloud NetApp Volumes\nperformance.\n\nBefore you begin\n----------------\n\nBefore you make changes to your volumes to optimize performance, review\n[performance considerations](/netapp/volumes/docs/performance/performance-considerations).\n\nAdjust volume settings\n----------------------\n\nYou can optimize performance by adjusting the following volume settings:\n\n- **Increase volume capacity**: You can increase the capacity of your Premium,\n Extreme or Standard service level volume to improve maximum achievable volume\n throughput. For volumes of Flex service level, increase storage pools capacity\n instead.\n\n- **Upgrade your service level**: You can upgrade your Premium service level\n volumes to the Extreme service level to improve throughput. We recommend that\n you assign the volume to a different storage pool with a different service\n level.\n\nIncreasing volume capacity and upgrading service levels are both non-disruptive\nto I/O workloads in process on the volume and don't affect access to the volume\nin any way.\n\nAdjust the client\n-----------------\n\nYou can improve performance by adjusting the following settings on the client:\n\n- **Co-locate clients**: latency results are directly impacted by the\n capabilities and location of the client. For best results, place the client\n in the same region as the volume or as close as possible. Test the zonal\n impact by testing latency from a client in each zone and use the zone with\n the lowest latency.\n\n- **Configure Compute Engine network bandwidth** : the network capabilities of\n Compute Engine virtual machines depend on the instance type used. Typically,\n larger instances can drive more network throughput. We recommend that you\n select a client virtual machine with an appropriate network bandwidth\n capability, select the Google Virtual NIC (gVNIC) network interface and\n enable `Tier_1` performance. For more information, see Compute Engine\n documentation on [network bandwidth](/compute/docs/network-bandwidth).\n\n- **Open multiple TCP sessions**: if your application requires high throughput,\n you can eventually saturate the single transmission control protocol (TCP)\n session that underlies a normal NFS and SMB session. For such cases, increase\n the number of TCP sessions your NFS and SMB connection uses.\n\n Use one of the following tabs to adjust your client based on the type of\n client: \n\n ### Linux\n\n Traditionally, an NFS client uses a single TCP session for all\n NFS-mounted file systems that share a storage endpoint. Using the\n [`nconnect` mount option](https://man7.org/linux/man-pages/man5/nfs.5.html)\n lets you increase the number of supported TCP sessions up to a maximum\n of 16.\n\n We recommend the following best practices for adjusting your Linux client\n type to fully take advantage of `nconnect`:\n - **Increase the number of TCP sessions with `nconnect`**: Each\n additional TCP session adds a queue for 128 outstanding requests,\n improving potential concurrency.\n\n - **Set `sunrpc.max_tcp_slot_table_entries` parameter** :\n `sunrpc.max_tcp_slot_table_entries` is a connection-level adjustment\n parameter which you can modify to control performance. We recommend\n setting `sunrpc.max_tpc_slot_table_enteries` to 128 requests or per\n connection and not surpassing 10,000 slots for all NFS clients within\n a single project connecting to NetApp Volumes. To set the\n `sunrpc.max_tcp_slot_table_entries` parameter, add the parameter to\n your `/etc/sysctl.conf` file and reload the parameter file using the\n `sysctl -p` command.\n\n - **Tune maximum supported value per session to 180** : Unlike NFSv3,\n NFSv4.1 clients define the relationship between the client and server\n in sessions. While NetApp Volumes supports up to 128\n outstanding requests per connection using NFSv3, NFSv4.1 is limited to\n 180 outstanding requests per session. Linux NFSv4.1 clients default to\n `64 max_session_slots` per session but you can tune this value as\n needed. We recommend changing the maximum supported value per session\n to 180.\n\n To tune `max_session_slots`, create a configuration file under\n `/etc/modprobe.d`. Make sure that no quotation marks (\" \") appear\n inline. Otherwise, the option doesn't take effect. \n\n $ echo \"options nfs max_session_slots=180\" \u003e /etc/modprobe/d/nfsclient/conf\n $ reboot\n\n Use the systool -v -m nfs command to see the current maximum in use\n by the client. For the command to work, at least one NFSv4.1 mount\n must be in place.\n\n $ systool -v -v nfs\n {\n Module = \"nfs\"\n ...\n Parameters:\n ...\n Max_session_slots = \"63\" \u003c-\n ...\n }\n\n The following NFS `nconnect` comparison graph demonstrates the impact\n using the nconnect configuration can have on an NFS workload. This\n information was captured using Fio with the following settings:\n - 100% read workload\n\n - 8 KiB block size against a single volume\n\n - `n2-standard-32` virtual machine using Red Hat 9 OS\n\n - 6 TiB working set\n\n Using an `nconnect` value of 16 resulted in five times more performance\n than when it wasn't enabled.\n\n ### Windows\n\n For Windows-based clients, the client can use [SMB Multichannel](https://learn.microsoft.com/en-us/previous-versions/windows/it-pro/windows-server-2012-r2-and-2012/dn610980(v=ws.11))\n with Receive Side Scaling (RSS) to open multiple TCP connections. To\n achieve this configuration, your virtual machine must have an allocated\n network adapter that supports RSS. We recommend setting RSS to four or\n eight values, however, any value over one should increase throughput.\n\n The following graph displays the difference using the RSS configuration\n can have on an SMB workload. This information was captured using Fio with\n the following settings:\n - 100% read workload\n\n - 8 KiB block size against a single volume\n\n - Single `n2-standard-32` virtual machine running a Windows 2022 OS\n\n - 6 TiB working set\n\n Eight jobs were run with only the SMB client RSS option changing between\n test executions. Using RSS values of 4, 8, and 16 increased performance\n two-fold when compared to using a value of 1. Each RSS instance was run\n nine times with a `numjobs` parameter of 8. The `iodepth` parameter was\n increased by five each execution until maximum throughput was reached.\n\nWhat's next\n-----------\n\nRead about [storage pools](/netapp/volumes/docs/configure-and-use/storage-pools/overview)."]]