Open LoganDark opened 1 year ago
Okay, now it's slow enough that when I started the portmaster app, it said the core service was not running for a second before it finally opened—and it's been trying to load the network activity for over a minute...
And the only "warning" is just a false alarm that cannot be dismissed in any way
Hello, just another user here. Welcome to hell.
On windows PM drops the ball terribly. The slow-down is because it intermittently answers (tested it out via WSL2's dig
and windows' nslookup
) DNS requests. Might be a throttling thing, might be a bug, no clue. So your software (Discord, whatever) spends a ridiculous amount of time attempting to resolve a hostname to an IP, but once it does the traffic flows because it's not your internet, it's PM.
At least that's my working theory at the moment.
I've played around with the resolvers, pointed PM at my own pihole install, pretty much everything I could think of. At this point I'm using Cloudflare's DNS which seems to be ok-ish in terms of speed. Using a privacy enabled DNS resolver like dnswatch or something results in almost an annoying amount of wait time.
I am glad to see your install detects the right exe making the requests. On my end everything is garbled and PM is 100% useless. But I'm on windows 11. So, a word of warning if you're planning on upgrading to 11.
I am glad to see your install detects the right exe making the requests
Oh, it doesn't. It sometimes decides Discord is trying to connect to Hacker News, or random Parsec domains are being accessed under "Other Connections". Why does it randomly get confused? No idea.
a word of warning if you're planning on upgrading to 11.
I use Portmaster on another computer running 11 and it works great.
Oh, it doesn't. It sometimes decides Discord is trying to connect to Hacker News, or random Parsec domains are being accessed under "Other Connections". Why does it randomly get confused? No idea.
Devs have no clue yet either, I opened 3 different tickets that ultimately stem from that problem. But now I'm sad because I thought Windows 10 was working fine.
I use Portmaster on another computer running 11 and it works great.
Well that's cool! Do you have WSL on it? Anything special about it?
Well that's cool! Do you have WSL on it? Anything special about it?
Yes, I have WSL 2 on it I believe. Not sure about "special" unless a liberal debloating (i.e. uninstalling Cortana, Windows Defender etc.) counts. It's a fresh install, home upgraded to professional, because I didn't trust the OEM install
It's caused by portmaster having a long uptime on a system that has a very very very active network
I know because the portmaster interface (the api used by the electron app) will start getting veeeeery slow and constantly freeze and portmaster will have to be restarted, except it will refuse to restart and there will be 2 core processes running because the old one cannot shut down on its own for some reason, so I have to kill it in task manager before the new one will take over and start working properly and suddenly everything is super fast and snappy again.
This is annoying because the new core will have forgotten the most recent dns requests and think everything is connecting to ips, but if I wait for the stupid prompts to time out then the next time any app retries its request, portmaster will catch on properly.
I think this is just stupid log bloat with o(n) or worse algorithms. Portmaster needs to have an eviction strategy
Hey @LoganDark, thanks for reporting this.
We have both fixed the process attribution on v1.0.11 as well as improved performance of handling connections. You can switch to beta to get it now or wait until next week: https://wiki.safing.io/en/FAQ/SwitchReleaseChannel
When you get the update, please report back whether the issues are fixed on your side too.
improved performance of handling connections
Does latency no longer scale with the size of the connections log? The UI gets really really slow when there are lots of connections in the log, even though it is paginated. It would be nice to evict old entries or provide a manual clear without having to restart Portmaster (which will refuse to shutdown cleanly if the connection log is too large already).
fixed the process attribution on v1.0.11
Awesome, you regresssed it terribly :3
It's even worse now and I have to turn prompt notifications off completely because of how broken Portmaster just became.
Since Portmaster doesn't support downgrading, I guess I'm stuck like this. (This is why I generally don't update software that works...)
I've moved this to https://github.com/safing/portmaster/issues/1157 - please stay on topic.
@dhaavi sorry I replied to your comment with something that directly addressed something you said. it wasn't me intentionally going off-topic.
@LoganDark Yes, you are right. I should not have brought the other issue into here myself.
Does latency no longer scale with the size of the connections log?
Yes, for the handling of connections in the core service - if the fix works as intended. CPU usage should not spike with many connections anymore.
No, for displaying connections in the UI, as more data to search/render is also always more work. This does not mean there is no room for improvement in the future.
It would be nice to evict old entries or provide a manual clear without having to restart Portmaster
This is a good idea. I will look into this.
which will refuse to shutdown cleanly if the connection log is too large already).
Please open a new issue for this, if this is consistent behavior.
No, for displaying connections in the UI, as more data to search/render is also always more work. This does not mean there is no room for improvement in the future.
Why are you processing the entire log when pagination means you only need to process a few tens of entries each load? (+ however many for the graph, of course)
Should this be a separate issue?
So I tried to do a random google search and it looks like this issue is getting much much worse:
Debug information:
funny how there is no connection to www.google.com
anywhere near the time I actually tried to open that page, and it's still loading because Portmaster seems to have completely failed to pick it up. Also, there's nothing in prompts.
There's something very wrong here apparently
This is still happening very frequently to us on the latest version of Portmaster. We have no insight into the issue and no idea why this is happening, but we constantly have to retry requests 3–4 times or wait up to 10 seconds in some cases for things as simple as:
These requests are usually picked up by the Portmaster core after a few seconds, but in extreme cases (see our reply right above) Portmaster can somehow forget to allow a connection to open, despite there being rules that permit it & it never generating a prompt, or even an entry in Portmaster's network log (which should happen even for connections that are prompting).
I wonder if this is some sort of bottleneck in Windows or just something with Portmaster, but in either case it's extremely frustrating and hindering. Do you have any statements about how this is prioritized internally or what we can do to help you resolve the issue?
-Emily
Is there any profiling or tracing I can do on my system in order to figure out what is the cause of this issue? I can do system-wide ETW traces or take the system into a kernel debugger or something. Is there absolutely anything that would help?
This behavior is ongoing in beta 1.0.12 as well. For example I keep getting DNS_PROBE_POSSIBLE for Brave which I've set to allow by default. This makes my web browsing experience feel like I'm on dialup in the 90s.
Maybe v1.0.13 looks to have something that could possibly help ("Suggest using stale DNS cache when queries are slow") but switching to the staging branch and repeatedly restarting and checking for updates does not find this version or expose this setting (even if I set the feature level to "Experimental") so I can't do any testing with this even though it was supposedly released 11 hours ago.
Has this been fixed? Is this why it's slow to solve DNS requests on my PC?
Has this been fixed? Is this why it's slow to solve DNS requests on my PC?
This has not been fixed. The recently added setting only allows connections to be identified by IP before a domain name has been resolved, which leads to false connection prompts
there is a new dns cashing option, Portmaster does prompt you to enable this setting if it detects slow dns speeds
there is a new dns cashing option, Portmaster does prompt you to enable this setting if it detects slow dns speeds
that is what I was just talking about, and is the setting that does not fix this issue. connections are still just as slow, because it is not caused by slow DNS
there is a new dns cashing option, Portmaster does prompt you to enable this setting if it detects slow dns speeds
I didn't notice any differences when enabling that option. I'm not sure how caching works, but wouldn't that speed up only websites that I already visited before and that therefore are cached? (I don't notice any difference even in that case though)
I don't think that this is a bug with Portmaster, it might be some compatibility thing then - this is always hard to tell
This seems to effect me even on a completely blank windows install.
You can easily test this by doing an nslookup from a client machine. With portmaster enabled, dns queries are consistently slow every time. Disable portmaster and responses are immediate with no lag whatsoever. This happens on all machines I have portmaster on, and like Bennett mentioned, also happens on brand new fresh Windows installs.
I don't think that this is a bug with Portmaster, it might be some compatibility thing then - this is always hard to tell
Have you tried testing it out in a simple VM? It happens on a fresh windows 10 or 11 install for me. It's very slow no matter what DNS config I give portmaster, including leave it blank to fallback to system DNS setting.
Even if I run portmaster with filtering disabled it still slows down all network requests. Sometimes my browser will hang for a long time and I have to blast it with refresh requests or open 8-15 tabs and paste the same URL in each before it finally resolves the domain and starts loading the content. Quite a terrible experience.
Sometimes my browser will hang for a long time and I have to blast it with refresh requests or open 8-15 tabs and paste the same URL in each before it finally resolves the domain and starts loading the content. Quite a terrible experience.
Absolutely, I hate this. I've gotten into the habit of always copying my web searches to the clipboard before I try to load them, so that I can paste it into a bunch more tabs if Portmaster decides it doesn't want to let that tab load. It's insane and probably counts as OCD at this point.
It consistently hangs on DNS resolution for me, testing in a loop:
while($val -ne 99)
{
$val++
(Measure-Command {Resolve-DnsName -Name www.google.co.uk -Server 1.1.1.1 -DnsOnly -NoHostsFile -Type A_AAAA}).TotalMilliseconds
}
Output:
83.6725
8.6989
14.6753
15.0204
14.4081
14.8532
14.6529
14.6084
14.9404
15.1601
14.9396
14.7747
14.4951
14.8335
14.3526
15.1252
14.9692
14.4629
14.6714
15.032
14.5034
14.7772
14.916
14.7585
14.6859
14.9887
15.0781
14.5636
14.9179
14.8247
12030.3671
81.3203
11.3967
14.897
14.792
15.0538
14.9636
14.889
14.975
15.43
15.078
15.1323
14.1221
14.4624
15.2542
12037.7361
71.3154
20.9707
15.093
14.5692
14.2499
15.1195
14.4399
14.9111
14.8097
15.1166
14.9051
14.3898
14.7166
14.8145
12049.5441
73.2265
18.6538
14.2874
15.2786
14.3787
14.9612
12060.2519
74.7417
18.1699
14.5706
14.8683
14.4668
14.7295
14.2383
14.9864
12050.3625
74.2015
18.4688
14.8863
14.7699
14.4545
14.7805
14.6655
14.7606
14.5938
14.3294
15.0409
14.9751
15.1185
15.5386
14.4999
14.6126
14.5667
14.6811
12047.0915
3101.1452
14.1336
12043.2231
Host is a windows 10 desktop on cable connection. For comparison, running the same loop in a macos on wifi at the same time has no timeouts.
Here is the output with Portmaster completely disabled:
I seem to be experiencing this too. I love the Portmaster software and its mission, but I simply, and regretfully, cannot use it with this bug around.
I have attempted three completely fresh installs of Portmaster now, and all three times it had this issue.
Without Portmaster, DNS requests take within 100 ms. From the moment I turn on Portmaster, my DNS requests are slowed down dramatically. Every request, no matter what, takes at least 700ms. Sometimes (relatively frequently) they won't resolve at all and will simply time out after hanging for several seconds (up to 10). Again, this happens pretty frequently. But even when they don't outright fail, all of my requests jump from sub-100 ms to 700+ ms.
This happens with completely clean installs, but I have also tried changing settings too. I've tried several different DNS settings, including removing them so it uses the system resolver. I've turned on caching, but there was no difference whatsoever. I'm not using a VPN, and I'm not using the SPN, but I also tried following all of the compatibility guides on the wiki, just in case; alas, none of them made any difference.
I've tried tweaking several other settings, turning off filters, everything I could think of. But nothing made any difference.
Once I turn off Portmaster however, my DNS requests immediately go back to normal and resolve at normal speed. If I turn Portmaster back on, they're back to slow again. And if I turn it off, they once again return to normal.
I really, really want to use Portmaster. But I cannot do it when not only is my basic computer usage dramatically slowed down, but my requests consistently fail. It is impossible to have an even passable experience doing something as simple as browsing the web when web pages consistently fail to load.
I truly hope the cause of this issue can be identified and resolved.
I highly doubt this is a "community helps community" situation because this seems to be an issue with portmaster that unconditionally degrades network performance in all cases which should make it a pretty severe, high priority issue
@Raphty any way for us to debug the delay? Maybe there are some logs or debug option on the app to figure out the resolution timeouts?
Delays are most likely coming from the process detection. We are currently working on an improvement there and I would wait for that before diving deeper into this issue.
We are currently working on an improvement there and I would wait for that before diving deeper into this issue.
This has been the case multiple times throughout the evolution of this problem, what makes this time different?
I'm writing this comment to let people know this has been an issue for me for the entire 8 months I've had Portmaster installed as well. Same as described in this thread, on Windows 10. I finally decided to either get the issue fixed, or find an alternative to Portmaster.
I'll be honest, I'm a little disappointed by the guy saying this does not seem to be an issue with Portmaster. Otherwise, Portmaster was great, but this is a dealbreaker.
I switched to pihole a few months ago because of this
This has been the case multiple times throughout the evolution of this problem, what makes this time different?
Not sure what we were working on the last times, but with v1.2 (currently in beta as v1.3), we improved the integration with the system to improve attributing connections. This means that the 5-50ms that were needed to attribute the connection are now pretty much gone.
(If you are on linux, update to v1.3, then wait 10 seconds, then reboot, and then check the debug info to see if there are any errors. If you have your own systemd unit file, you might need to update that yourself to allow eBPF usage.)
If this issue continues after v1.3 I will take another close look into this.
Also, in most cases this will probably be DNS. So please check your configured DNS server in Portmaster, or try remove all DNS servers from the config and fallback to the system DNS to test if that is slow for you.
I tried v1.3 of Portmaster on Windows with system default DNS. There is some degree of general slowdown by just enabling Portmaster, but I think it is at an acceptable level.
What is not acceptable however, and what got me to reluctantly uninstall Portmaster once again as I wait for a future fix, is that all of the other issues I had before still remain.
The first issue is that whenever I start my PC, it seems like connections will just refuse to go through for a minute or more.
If I try to go to websites in the browser, either the browser will just pretend like I didn't enter anything, or it will say DNS lookup failed. It will do this for every request for about a minute or two before requests can even begin to go through.
I'd like to reiterate that:
This issue seems related to what Working-Name and LoganDark experienced from earlier in this thread.
The second issue is the same one that lkraider posted about earlier in this thread.
When I have Portmaster enabled, requests will VERY FREQUENTLY simply hang for an excruciatingly long time. As you can see in lkraider's post, his requests frequently take 12000+ ms to resolve. This is the same behavior I am experiencing. If I run the exact same command lkraider did, I'll see it full of requests that take several thousand—often over 10,000—ms to resolve. Sometimes they will take so long that the command says the connection simply timed out. It wont be every request, just like lkraider, but it will be several of them.
I would once again like to clarify that these issues have been going on for me for several versions. Each time I attempt to install Portmaster, and each time these issues make me uninstall. They continue to be issues on v1.3.
Hey @Andell4301, thanks for the thorough report and summary of issues.
I will try to reproduce this in the coming weeks and will report back with my findings.
Issue 2
The second issue is the same one that lkraider posted about earlier in this thread.
And me too, FWIW. I even have a video of Google Search completely refusing to load until Portmaster is restarted in order to allow the connection through. This happens all the time regardless of how recently I started my computer, browser, or Portmaster.
One thing that would be very interesting to know, is if disabling the Seamless DNS Integration setting changes anything.
You'll need to dig out this setting and you should not use it in production, as it makes Portmaster partly blind and renders a lot of features useless. But as this disables mangling of DNS records, it could point us in the right direction to debug this.
So, if anyone who is experiencing this and can try and report back, that'd be super helpful! Thanks!
One thing that would be very interesting to know, is if disabling the Seamless DNS Integration setting changes anything.
while scrolling down to find that setting, I saw this and actually I think I'm done troubleshooting portmaster for today (mistake? or time to update your marketing that claims the subscription is only to cover SPN costs?)
There is a change coming here, yes. We've talked about it in the last two live streams and the pricing page is already updated: https://safing.io/pricing/ There will be a more extensive blog post next week going into the details.
(Marking the last and this comment as off-topic to stay focused here. If there are any questions, please ask on discord.)
I think it's a huge mistake to charge for a faulty, unfinished product - SPN or not. You will have successfully shot yourself in the foot by asking for money for something you cannot deliver for 1+ years.
Handle the bugs, make it do what is advertised, then charge. And when you do, maybe don't pull a Teamviewer, LogMeIn, Dropbox, 1Password, or whatever other bait&switch and put previously available features behind a paywall. Nothing screams untrustworthy more, and unfortunately for you, you're in the business of security/privacy.
I think it's a huge mistake to charge for a faulty, unfinished product - SPN or not. You will have successfully shot yourself in the foot by asking for money for something you cannot deliver for 1+ years.
Handle the bugs, make it do what is advertised, then charge. And when you do, maybe don't pull a Teamviewer, LogMeIn, Dropbox, 1Password, or whatever other bait&switch and put previously available features behind a paywall. Nothing screams untrustworthy more, and unfortunately for you, you're in the business of security/privacy.
SPN isn't faulty and a SaaS is never finished. But suddenly beginning to restrict which local, on-machine analysis I am allowed to perform based on a subscription that used to represent only access to paid servers, not access to paywalled features on my own computer is definitely a change that will take some getting used to. Especially since Portmaster even used to keep track of long-term logs, but then they nerfed it (to be fair, it was causing a performance problem back then).
That history analysis feature on its own, FWIW, is completely new, so they're not locking down existing features at all. Just gating new ones.
No, Portmaster is faulty on windows. I never tried SPN, because it has no value proposition for me.
Portmaster hasn't worked normally on Windows ever since I first tried it 1+ years ago. And the direction is backwards (UI has become more cumbersome, not more user friendly), with lots of pushback from the devs - oh, it's Windows' fault for this and that. Oh, it's your fault because you cleared your DNS entries... anything but the actual software.
To be fair, though, "connection history" sounds like what PM has done all along, I didn't get the distinction of on-machine analysis.
For example, where's the "live view" for connections? You have to hit refresh every time you want to see the new connections being made by an app.
No, Portmaster is faulty on windows. I never tried SPN, because it has no value proposition for me.
Portmaster hasn't worked normally on Windows ever since I first tried it 1+ years ago. And the direction is backwards (UI has become more cumbersome, not more user friendly), with lots of pushback from the devs - oh, it's Windows' fault for this and that. Oh, it's your fault because you cleared your DNS entries... anything but the actual software.
To be fair, though, "connection history" sounds like what PM has done all along, I didn't get the distinction of on-machine analysis.
For example, where's the "live view" for connections? You have to hit refresh every time you want to see the new connections being made by an app.
Portmaster is currently the most technically advanced firewall available for Windows. Say what you will about the user interface and feature set, but unless you want something that MitMs encrypted connections, nothing else out there has per-process, per-domain, per-port, per-protocol control over your entire computer, including all privileged operating system processes. Not to mention the ability to give you live prompts to allow selected first-time connections to succeed—I've seen live prompts implemented in other ways before, but Portmaster (similarly to Little Snitch for macOS, which is where I came from) is the only firewall that can prompt me about a connection without actually blocking it first.
(FWIW, I also have never paid for Portmaster and have never used the SPN.)
What happened:
Sometimes, when sending an HTTP request or opening some other kind of connection, it will just... hang for 5-10 seconds. It's not my internet or the other server, and Portmaster correctly logs and attributes the connection to the right process, it's just... some things are slow. And this is for allowed connections. Like, websites will take a long time to open but will then load quickly, and Discord messages will occasionally take way too long to send, among other random things throughout the system.
What did you expect to happen?:
Portmaster shouldn't delay any allowed connections
How did you reproduce it?:
I have Windows Defender Firewall fully disabled, and I'm using the network over a USB-3-to-Ethernet dongle, if that matters.
I have Portmaster set to prompt for connections that don't have associated rules, but.... these connections are not being prompted. They just... take a few seconds to open even though they are already allowed.
Debug Information:
Version 1.0.7
``` Portmaster version 1.0.7 commit tags/v1.0.7-0-gdc5dd359bfef1e739ef06fd439d1df15e3603c7c built with go1.19 (gc) windows/amd64 using options main.go by user@docker on 21.02.2023 Licensed under the AGPLv3 license. The source code is available here: https://github.com/safing/portmaster ```Platform: Microsoft Windows 10 Home 10.0.19041 Build 19041
``` System: Microsoft Windows 10 Home windows (Standalone Workstation) 10.0.19041 Build 19041 Kernel: 10.0.19041 Build 19041 x86_64 ```Status: Trusted
``` ActiveSecurityLevel: Trusted SelectedSecurityLevel: Off ThreatMitigationLevel: Trusted CaptivePortal: OnlineStatus: Online ```Config: 11
``` core/automaticUpdates: false core/expertiseLevel: developer dns/nameserverRetryRate: 30 dns/nameservers: [redacted] dns/noAssignedNameservers: 7 dns/noInsecureProtocols: 7 filter/blockInbound: 6 filter/defaultAction: ask filter/endpoints: [redacted] filter/lists: [TRAC MAL DECEP BAD CB-MW WSP SH-CRL SH-OCSP UNBREAK] spn/use: false ```Resolvers: 10/10
``` Cloudflare (dot://cloudflare-dns.com:853#config) dot://cloudflare-dns.com:853#config Failing: false Cloudflare (dot://cloudflare-dns.com:853#config) dot://cloudflare-dns.com:853#config Failing: false 2606:4700:4700::1111 (dns://2606:4700:4700::1111:53#system) dns://2606:4700:4700::1111:53#system Failing: false 2606:4700:4700::1001 (dns://2606:4700:4700::1001:53#system) dns://2606:4700:4700::1001:53#system Failing: false 2001:4860:4860::8888 (dns://2001:4860:4860::8888:53#system) dns://2001:4860:4860::8888:53#system Failing: false 2001:4860:4860::8844 (dns://2001:4860:4860::8844:53#system) dns://2001:4860:4860::8844:53#system Failing: false 1.1.1.1 (dns://1.1.1.1:53#system) dns://1.1.1.1:53#system Failing: false 1.0.0.1 (dns://1.0.0.1:53#system) dns://1.0.0.1:53#system Failing: false 8.8.8.8 (dns://8.8.8.8:53#system) dns://8.8.8.8:53#system Failing: false 8.8.4.4 (dns://8.8.4.4:53#system) dns://8.8.4.4:53#system Failing: false ```SPN: disabled (module disabled)
``` HomeHubID: HomeHubName: HomeHubIP: Transport: --- Client: true PublicHub: false HubHasIPv4: false HubHasIPv6: false ```Compatibility: WFP State (19)
``` Edge traversal Teredo Authorization Sublayer SubLayer Edge traversal Teredo Authorization Sublayer {7b6b11f6-cbb5-433c-ae06-6a4f0076e49e} IPxlat Forward IPv4 filter Callout Filters forwarded IPv4 packets into synthetic IPv6 packets {b255c296-7e0c-4115-95f3-b7f24a8a1162} [no provider key] FWPM_LAYER_IPFORWARD_V4 IPxlat Forward IPv4 sub layer SubLayer Sub layer for filtering forwarded IPv4 packets into synthetic IPv6 packets {4351e497-5d8b-46bc-86d9-abccdb868d6d} IPxlat Inbound IPv6 filter Callout Filters incoming IPv6 packets into synthetic IPv4 packets {93bb703d-0502-42e2-8e30-a14576e5085d} [no provider key] FWPM_LAYER_INBOUND_IPPACKET_V6 IPxlat Inbound IPv6 sub layer SubLayer Sub layer for filtering incoming IPv6 packets into synthetic IPv4 packets {dfb035ca-c2a7-4684-97b6-4dbc57c63590} IPxlat Outbound IPv4 filter Callout Filters outgoing IPv4 packets into synthetic IPv6 packets {66d52657-1979-4e58-b3f7-4756434c4880} [no provider key] FWPM_LAYER_OUTBOUND_IPPACKET_V4 IPxlat Outbound IPv4 sub layer SubLayer Sub layer for filtering outgoing IPv4 packets into synthetic IPv6 packets {d3e70856-fc90-4c0a-b9b2-a6f73e20b5cc} PortmasterInboundV4Callout Callout This callout is used by the Portmaster to intercept inbound IPv4 traffic. {05c55149-4732-4857-8d10-f178f3a06f8c} [no provider key] FWPM_LAYER_INBOUND_IPPACKET_V4 PortmasterInboundV4Filter Filter This filter is used by the Portmaster to intercept inbound IPv4 traffic. {4d24331c-a656-4b98-9d9f-266eff2a533e} [no provider key] FWPM_LAYER_INBOUND_IPPACKET_V4 {a87fb472-fc68-4805-8559-c6ae774773e0} PortmasterInboundV6Callout Callout This callout is used by the Portmaster to intercept inbound IPv6 traffic. {ceff1df7-2baa-44c5-a6e5-73a95849bcff} [no provider key] FWPM_LAYER_INBOUND_IPPACKET_V6 PortmasterInboundV6Filter Filter This filter is used by the Portmaster to intercept inbound IPv6 traffic. {006c4a4a-3066-45ce-8360-90c48555c66b} [no provider key] FWPM_LAYER_INBOUND_IPPACKET_V6 {a87fb472-fc68-4805-8559-c6ae774773e0} PortmasterOutboundV4Callout Callout This callout is used by the Portmaster to intercept outbound IPv4 traffic. {41162b9e-8473-4b88-a5eb-04cf1d276b06} [no provider key] FWPM_LAYER_OUTBOUND_IPPACKET_V4 PortmasterOutboundV4Filter Filter This filter is used by the Portmaster to intercept outbound IPv4 traffic. {b69fe9ab-6ca2-422a-a246-41367aa04a06} [no provider key] FWPM_LAYER_OUTBOUND_IPPACKET_V4 {a87fb472-fc68-4805-8559-c6ae774773e0} PortmasterOutboundV6Callout Callout This callout is used by the Portmaster to intercept outbound IPv6 traffic. {32bad112-6af4-4109-809b-c07570ba01b4} [no provider key] FWPM_LAYER_OUTBOUND_IPPACKET_V6 PortmasterOutboundV6Filter Filter This filter is used by the Portmaster to intercept outbound IPv6 traffic. {8888fde4-e98b-4133-ab83-8d5b45a43405} [no provider key] FWPM_LAYER_OUTBOUND_IPPACKET_V6 {a87fb472-fc68-4805-8559-c6ae774773e0} PortmasterSublayer SubLayer The Portmaster sublayer holds all it's filters. {a87fb472-fc68-4805-8559-c6ae774773e0} Private Internet Access Firewall Provider Implements privacy filtering features of Private Internet Access. {08de3850-a416-4c47-b3ad-657c5ef140fb} Private Internet Access Firewall SubLayer Implements privacy filtering features of Private Internet Access. {f31e288d-de5a-4522-9458-de14ebd0a3f8} Teredo socket option opt out block filter Filter [no description] {452c2f19-bef9-4738-bcf5-04f97ee5f04e} {3b4cc995-4067-4d73-914c-31c2ccf09530} FWPM_LAYER_ALE_AUTH_RECV_ACCEPT_V6 {7b6b11f6-cbb5-433c-ae06-6a4f0076e49e} ```Notifications Module Error
``` Message: panic: runtime error: invalid memory address or nil pointer dereference Timestamp: 2023-03-21 10:46:22.9235129 -0700 PDT m=+748805.290658501 ModuleName: notifications TaskName: notification action execution TaskType: worker Severity: panic PanicValue: runtime error: invalid memory address or nil pointer dereference StackTrace: goroutine 1314617 [running]: runtime/debug.Stack() /usr/local/go/src/runtime/debug/stack.go:24 +0x65 github.com/safing/portbase/modules.(*Module).NewPanicError(0xc00018a8c0, {0xb27fc0, 0x1d}, {0xaf5879, 0x6}, {0xa10f40?, 0x1254ec0}) /home/user/git/safing/portbase/modules/error.go:61 +0x94 github.com/safing/portbase/modules.(*Module).runWorker.func1() /home/user/git/safing/portbase/modules/worker.go:121 +0x71 panic({0xa10f40, 0x1254ec0}) /usr/local/go/src/runtime/panic.go:884 +0x212 github.com/safing/portmaster/profile.GetLocalProfile({0xc0026446f0, 0x24}, {0x0, 0x0}, 0x0) /home/user/git/safing/portmaster/profile/get.go:133 +0xa27 github.com/safing/portmaster/firewall.saveResponse(0xc0007d21e0?, 0xc001973a20, {0xc000524bb0, 0x10}) /home/user/git/safing/portmaster/firewall/prompt.go:263 +0x145 github.com/safing/portmaster/firewall.createPrompt.func1({0xc0022ad0e0?, 0xc0001b5680?}, 0xc00113fe88?) /home/user/git/safing/portmaster/firewall/prompt.go:191 +0x32 github.com/safing/portbase/notifications.(*Notification).selectAndExecuteAction.func1({0xc9b1b8?, 0xc000731900?}) /home/user/git/safing/portbase/notifications/notification.go:461 +0x2e github.com/safing/portbase/modules.(*Module).runWorker(0x0?, {0xb27fc0?, 0xca4290?}, 0xc000709720?) /home/user/git/safing/portbase/modules/worker.go:130 +0x93 github.com/safing/portbase/modules.(*Module).RunWorker(0x1746b7?, {0xb27fc0?, 0xc0007095e0?}, 0xc00113ffb8?) /home/user/git/safing/portbase/modules/worker.go:52 +0x77 github.com/safing/portbase/modules.(*Module).StartWorker.func1() /home/user/git/safing/portbase/modules/worker.go:27 +0x4a created by github.com/safing/portbase/modules.(*Module).StartWorker /home/user/git/safing/portbase/modules/worker.go:26 +0xac ```Unexpected Logs
``` 230321 09:30:26.418 ate/tables:037 > WARN 113 state: failed to get UDP6 socket table: insufficient buffer error (tried 5 times): provided 1024 bytes; required 1328 bytes - [NT 0x7A] The operation completed successfully. ```Updates: stable (6/29)
``` Active: all/intel/geoip/geoipv4.mmdb.gz: 20230102.9.32 all/intel/geoip/geoipv6.mmdb.gz: 20230102.13.14 all/intel/portmaster/notifications.yaml: 20230227.8.35 all/ui/modules/assets.zip: 0.3.1 all/ui/modules/portmaster.zip: 0.4.5 windows_amd64/kext/portmaster-kext.sys: 1.0.17 Selected: all/intel/geoip/geoipv4.mmdb.gz: 20230102.9.32 all/intel/geoip/geoipv4.mmdb: 20230102.9.32 all/intel/geoip/geoipv6.mmdb.gz: 20230102.13.14 all/intel/geoip/geoipv6.mmdb: 20230102.13.14 all/intel/lists/base.dsdl: 20230301.0.3 all/intel/lists/index.dsd: 2022.6.7 all/intel/lists/intermediate.dsdl: 20230305.0.1 all/intel/lists/urgent.dsdl: 20230311.15.1 all/intel/portmaster/notifications.yaml: 20230227.8.35 all/intel/spn/main-intel.yaml: 20230310.10.36 all/ui/modules/assets.zip: 0.3.1 all/ui/modules/base.zip: 0.2.11 all/ui/modules/console.zip: 0.1.11 all/ui/modules/monitor.zip: 0.2.4 all/ui/modules/portmaster.zip: 0.4.5 all/ui/modules/profilemgr.zip: 0.1.7 all/ui/modules/settings.zip: 0.1.8 windows_amd64/app/portmaster-app.zip: 0.2.5 windows_amd64/core/portmaster-core.exe: 1.0.7 windows_amd64/hub/spn-hub.exe: 0.6.2 windows_amd64/jess/jess.exe: 0.3.1 windows_amd64/kext/portmaster-kext.dll: 1.0.14 windows_amd64/kext/portmaster-kext.pdb: 1.0.17 windows_amd64/kext/portmaster-kext.sys: 1.0.17 windows_amd64/notifier/portmaster-notifier.exe: 0.3.5 windows_amd64/notifier/portmaster-snoretoast.exe: 0.6.0 windows_amd64/notifier/portmaster-wintoast.dll: 0.1.4 windows_amd64/packages/portmaster-installer.exe: 1.0.0 windows_amd64/start/portmaster-start.exe: 1.0.2 ```Goroutine Stack
!["Comment is too long (maximum is 65536 characters)", "There was an error creating your Issue: body is too long (maximum is 65536 characters)."](https://user-images.githubusercontent.com/4723091/226697837-77e677bf-f919-496f-bd3f-ec590cf7e434.png)