At a glance

If you manage two or more subscriptions in v2rayN, create a group for each provider and make sure each one updates on its own before adding node-name filters. Then compare latency and speed within each group. Once you’ve confirmed the new list works, enable automatic updates. If a provider changes its naming scheme, you can troubleshoot just that group instead of starting over.

Create a group for each subscription and update it once without filters

Here, “provider” means a service that supplies subscription links; a subscription URL is the link the client uses to fetch its server list. Create one group per subscription and include both the source and purpose in its name, such as “Provider A - Daily” or “Provider B - Backup.” Avoid grouping by region first: both providers might have a node called “Hong Kong 01,” making it hard to tell which subscription to update.

Subscription groupsFetch server listMatch namesTest within groupConfirm candidates
  1. Using the v2rayN 7.x interface on Windows as a reference, open “Subscription Group” → “Subscription Group Settings,” click “Add,” then enter a remark and its subscription URL. Create a separate group for each provider; don’t paste two URLs into the same field.
  2. Leave “Server Filter” empty when you first save. Return to the main window, select the new group, then use the “Subscription Group” menu to update the current subscription. Menu labels may vary slightly between minor versions. Make sure you choose the current subscription, not all subscriptions.
  3. After the refresh, note the number of nodes in the group and a few representative remarks: ones to keep, ones to exclude, and ones that could be easy to misread. If the list is empty, check the update log, subscription URL, and server response first. Since no filter is set yet, don’t start by changing the regex.

How to set a server filter: use regex to keep matching nodes

“Server Filter” uses a regular expression to match the remarks on imported nodes—the names shown in the list. A regular expression is a way to describe a text-matching rule. When you add a rule, matching entries are kept during updates. It doesn’t delete nodes containing a few words, or detect a server’s actual location, test results, or whether its protocol works.

Select the relevant group, then edit its “Server Filter” in “Subscription Group Settings.” For example, to keep only nodes whose names start with “Hong Kong” or “Japan,” start with the rule below. ^ anchors the match to the start of the name, | means either option, \s matches whitespace, and $ marks the end of the name.

^(Hong Kong|Japan)(\s|$)
Node remarkResultWhy
Hong Kong 01KeepThe region is at the start, followed by a space.
JapanKeepThe region is at the start, with nothing after it.
Japan-02ExcludeA hyphen isn’t one of the separators specified by this rule.
Premium Hong Kong 03Exclude“Hong Kong” isn’t at the start of the name.

Should both providers use the same rule?

Daily group: full region names

Subscription remark
Provider A - Daily
Example names
Hong Kong 01, Japan 02
Server filter
^(Hong Kong|Japan)(\s|$)

If the names use hyphens, replace the ending (\s|$) with ([\s-]|$), then update and check the results.

Backup group: region codes

Subscription remark
Provider B - Backup
Example names
SG-01、JP-02
Server filter
^(SG|JP)-[0-9]+$

This rule keeps only names with a region code followed by a number, so JP-02-B won’t match.

After saving, update the current subscription again. Don’t judge the match against the old list from before you saved the rule. If the provider adds a flag or batch number before the region, an old rule with ^ may match nothing. Clear the filter for this group and update it again to check the new remarks before editing the rule. Don’t change the URL and the rule at the same time, or it’ll be hard to tell what caused the problem.

Test within each group: check connectivity first, then compare download speeds

After filtering, switch to the group you want and select a few candidate servers. From the list’s right-click menu, run a real connection latency test first, then speed-test a smaller set of candidates that passed. Real connection latency measures how long that node takes to reach the test target; it doesn’t represent response times for every website. Speed tests also use subscription traffic, so there’s no need to test every node after every update.

These sample readings from the same group and time period illustrate how to choose a node; they aren’t actual benchmark results. Assume both routes can open the sites you normally use and were tested against the same target:

92 ms
Hong Kong 01 real connection latency
4.8 MB/s
Hong Kong 01 download speed
148 ms
Japan 02 real connection latency
9.2 MB/s
Japan 02 download speed

Takeaway: test everyday browsing and large downloads separately

For everyday browsing, try Hong Kong 01 first because it has lower latency. If you often download large files, retest Japan 02 with a file you actually need to download. If you get an occasional timeout, run two more tests on the same network before deciding whether to move it to your backup group.

  • When comparing groups, use the same device, network, and test target, and test them as close together in time as possible. Don’t compare daytime results directly with results from the evening peak.
  • If every node in a group fails, try reaching the same test target with a known-working node first. If the baseline node fails too, check your local network or the test target before deleting the entire group.
  • After testing, choose your everyday node manually. Selecting a subscription group only displays that group’s list; it doesn’t switch the active connection. Updating a subscription doesn’t automatically switch to the fastest node, either.

Set automatic update intervals and keep a manual recovery option

Once both groups have updated manually and the filters work as expected, go to “Subscription Group” → “Subscription Group Settings” and set an automatic update interval for each group. If the interface uses minutes, 1440 is 24 hours and 4320 is 72 hours. Try 1440 minutes for the daily group and 4320 for the backup group. If your provider requires a longer refresh interval, follow its guidance.

  1. Check the log after an update is due. Automatic updates require the client to be running; they aren’t a background service when your computer is off or the client is closed. Check the group’s last update time and log instead of relying on the node count to tell you whether it refreshed.
  2. If an update fails, check how the URL is reached. First, try updating the current group without using a proxy. If the subscription URL is only accessible through a working proxy, try the update option that uses a proxy. A failed subscription fetch and a failed node connection are different problems; identify which step failed first.
  3. If an update succeeds but the filter behaves unexpectedly, restore the full list first. Temporarily disable automatic updates for the affected group, save the current rule, clear the filter, and manually update that group. If the list comes back, rewrite the rule to match the new remarks. If it still doesn’t, check whether the provider changed the URL or returned content that can’t be imported.
  4. Make sure the current node still works. Providers may rename, replace, or remove nodes in a subscription. If the node you selected no longer works after an update, manually choose a tested candidate and confirm it by opening a site you use regularly. Then re-enable automatic updates for that group.