Wonder if any old software is going to have an issue today
lol
@ejsmith @pwelter34 how is logging coming along? We really should update https://github.com/exceptionless/Foundatio.Samples and https://github.com/exceptionless/Foundatio.Repositories as well
you guys have been fixing some good bugs and the only way we can get them is by upgrading
applied 16 updates, a free cpu upgrade and complete reboot in under 3 minutes
thing is so insanely fast
32$ a moth for a quad core 3.6ghz cpu, 90gb ssd
I split it with a old developer friend
so it’s 16$ a month
brb
<script src="//cdn.rawgit.com/exceptionless/Exceptionless.JavaScript/v1.3.2/dist/exceptionless.min.js?apiKey=YOUR_API_KEY"></script>
<script src="//cdn.jsdelivr.net/fingerprintjs2/1.1.1/fingerprint2.min.js"></script>
<script src="//cdn.jsdelivr.net/fingerprintjs2/1.1.1/fingerprint2.min.js"></script>
<script type="text/javascript">
new Fingerprint2().get(function(uniqueHash, components) {
var client = exceptionless.ExceptionlessClient.default;
client.config.setUserIdentity(uniqueHash);
client.submitFeatureUsage('PageView');
});
</script>
wonder if I need to do on page load with that baby
that should give me unique anon users without cookies
I noticed that I had to go through the source to figure this out
so we have some work todo
Hmmm... seems a bit overkill. If we just store a cookie or an id in local storage, it would be better because you can use that and know it won't change unlike this fingerprint.
but until we have something like that
..
How hard is it to add an id somewhere? Do it in your own code.
Says that takes up to 600ms to run.
but then I have to have startup logic to persist it and get it
guess I can see how easy it is to do it
there is no way I can do that
@frankebersoll
our storage settings is queue based
and it’s really hard to get a stored key out of it
private static getSavedServerSettings(config: Configuration): ISettingsWithVersion {
let item = config.storage.settings.get()[0];
if (item && item.value && item.value.version && item.value.settings) {
return item.value;
}
return { version: 0, settings: {} };}
before we could do config.storage.get(maxitems)
now we moved it to storage having different containers (.queue and .settings).
but no way to get named things out of them
so makes it really hard for me to use the storage myself
wonder what would be a good way around that
would be nice if we had scoped keys and a single collection..
that way queue items would be prefixed and settings likewise and then we could use a get(“namespace*”) or get(“id”)
we should have different defaults for settings storage and queue storage.
settings should default to being persisted.
and queue defaults to in memory.
the version needs to be fixed
I’m using it
what is broken?
the nuget package is broken
look at it
target 4.5 and try to use it
ok, so that doesn’t keep you from using it, but yes it should be fixed.
because our package targets 4.5
so nuget picks 4.5 instead of 4.5.1
and then won’t compile due to no reference
I think they should take a logger factory
and then we get can get rid of the enablelogging flag
want me to to make that change?
I’m in the middle of something
then going to lunch
I was just going to update this foundatio repos change quick (not updating anything else) just want to keep track of everything else that is going on)
didn’t you add something to es config to keep the amount of logs limited?
bunch of duplicated types
nope
remember some guy created an issue
I think the 1 time handlers are a WIP
not working
yeah there was an issue for them
but I didn’t apply it
because it really changes it up
the format
and it’s easy to wipe them out if we do have an issue
Seems like I can’t update those.. and they were brought in by the xunit project I believe.
InMemoryMetricsClient: An in memory metrics implementation. This metrics implementation is only valid for the lifetime of the process. It's worth noting that this metrics client also has the ability to display the metrics to a TextWriter on a timer or by calling DisplayStats(TextWriter).
where is display stats
got rid of it… we have the ability to view metric stats now.
should update the readme
// 008, Accoona, Arachmo, boitho, B-O-T, Cerberian Drtrs, Charlotte, Charlotte, Covario IDS, DataparkSearch,FindLinks,Holmes,htdig,
// ia_archiver,ichiro,igdeSpyder,L.webis,Larbin,LinkWalker,lwp-trivial,mabontland,Mediapartners-Google,Mnogosearch,mogimogi,Morning Paper,
// MVAClient,NetResearchServer,NewsGator,NG-Search,NutchCVS,Nymesis,oegp,Orbiter,Peew,Pompos,PostPost,PycURL,Qseero,Radian6,SBIder
// ScoutJet,Scrubby,SearchSight,semanticdiscovery,ShopWiki,silk,Snappy,Sqworm,StackRambler,Teoma,TinEye,truwoGPS,updated,Vagabondo,Vortex
//voyager,VYU2,webcollage,Websquash.com,wf84,WomlpeFactory,yacy,Yahoo,YandexImages,Yeti,yoogliFetchAgent,Zao, ZyBorg
project.Configuration.Settings[SettingsDictionary.KnownKeys.UserAgentBotPatterns] = "*bot*,*crawler*,*spider*";
comments are the ones not covered by the wildcards I’m going to start setting for new projects and in this update work item job that’s triggered via admin endpoint
see any in there you care about
or should I just add them all
tineye, newsgator and yahoo stand out to me personally
well, hopefully adding all of those won’t block any legit traffic.
since we use wildcards
unless you think we should add all of the above
I think if they are unique bot agents that won’t accidentally remove valid traffic...
then we should add them as defaults
just worried that some will be too generic
guess I can add all of those :\
just think that it’s a huge list
I think the three above and maybe yahoo
would cover like 95%
or higher
do the big ones
head
do you know if we are releasing locks if an error occurs inside of a job?
not sure
via using block
but I wonder
I triggered this job and hit an exception
and then my work item job isn’t being rehit
it has a default delay
I am seeing an issue where the queue job appears to stop working.
gotta try and figure that out.
whose going to be pull request #200?
@ejsmith exceptionless/Exceptionless#199
this fixes a few issues of which bill ran into and should also help with deduping
well.. noise..
@ejsmith are you going to be on tonight?
We can review my stats stuff
and other pull request stuff :)
@niemyjski why are you throttling the work item to once every 15 minutes?