Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 217 218 219 220 221 222 223 224 225 226 227 .. 5022
— 502,152 events total
2020-09-26 05:53:43 <dsal> I use influxdb from Haskell quite a bit. My coworkers like prometheus, but I don't respond to their peer pressure.
2020-09-26 05:53:43 × heatsink quits (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net) (Remote host closed the connection)
2020-09-26 05:53:45 × berberman quits (~berberman@2408:8207:256c:1c00:584e:a9ff:fe9b:d3fe) (Quit: ZNC 1.7.5 - https://znc.in)
2020-09-26 05:54:07 <Guest26> hmm, yes, certainly it is not supposed to be doing this, metrics thing
2020-09-26 05:54:08 → o1lo01ol1o joins (~o1lo01ol1@bl8-213-81.dsl.telepac.pt)
2020-09-26 05:54:19 <fraktor> Prometheus is not really great as a time-series database generally, but if you're doing metrics, then it's absolutely the tool (or a very good tool anyway) for the job.
2020-09-26 05:54:38 <Guest26> im not
2020-09-26 05:54:51 <Guest26> it was just a simple example of a time varying value
2020-09-26 05:55:01 × Tops2 quits (~Tobias@dyndsl-095-033-024-124.ewe-ip-backbone.de) (Quit: Leaving.)
2020-09-26 05:55:04 <fraktor> Okay. In that case, a time-series database with a Haskell backend is a good choice. I haven't used influxdb but I've heard good things.
2020-09-26 05:55:09 <Guest26> but a random number based on time of day is a better example
2020-09-26 05:55:13 <fraktor> If you don't mind me asking, what is the data that you're sending?
2020-09-26 05:55:25 → thir joins (~thir@p200300f27f0fc60094e773283d7bf825.dip0.t-ipconnect.de)
2020-09-26 05:55:32 × AmberJ_ quits (~AmberJ_@195.140.213.38) (Remote host closed the connection)
2020-09-26 05:56:03 <Guest26> like i say, generally any json serializable, or perhaps binary serializable data. so generally any type that supports a generic instance.
2020-09-26 05:56:29 <fraktor> Is it all time-series data?
2020-09-26 05:56:35 <EvanR> a single random number based on the current time of day can be pretty bad
2020-09-26 05:56:43 <dsal> Even 4?
2020-09-26 05:56:53 <EvanR> 4 is one of the worst
2020-09-26 05:56:53 <Guest26> but if it has to go to an SQL database so it can be read by the HTML page, then maybe this limits the type
2020-09-26 05:57:09 <fraktor> I don't mean to pry but, what is the problem you're trying to solve?
2020-09-26 05:57:23 × sand_dull quits (~theuser@185.217.69.182) (Ping timeout: 246 seconds)
2020-09-26 05:57:30 <Guest26> fraktor: its not time series data persay, refreshing the webpage should just visualise the current database
2020-09-26 05:57:47 <Guest26> it would be time series data if the database update were just simple addition of a new value
2020-09-26 05:57:48 <dsal> What is "the current database" ?
2020-09-26 05:58:05 × o1lo01ol1o quits (~o1lo01ol1@bl8-213-81.dsl.telepac.pt) (Ping timeout: 240 seconds)
2020-09-26 05:58:17 <Guest26> thats quite arbitrary - but for the sake of argument its a list of (time of day,double)
2020-09-26 05:58:41 <fraktor> You can absolutely use Haskell as a backend for some kind of frontend, and that would solve that problem, but what on earth are you doing with it?
2020-09-26 05:58:43 <Guest26> at least then i can imagine a simple HTML based visualisation thats just an xy graph
2020-09-26 05:58:49 <fraktor> Like, what is the software you're writing trying to solve?
2020-09-26 05:58:55 <dsal> Is a database even helpful? If your goal is to put all the data in a web page, then what does the database do here?
2020-09-26 05:59:19 <fraktor> dsal: I think that they were saying they didn't want to use the database if possible
2020-09-26 05:59:19 <Guest26> the database was for when i was thinking to visualise the data using R
2020-09-26 05:59:38 <Guest26> the R tutorial here used an SQL server; https://www.r-bloggers.com/2018/07/real-time-data-visualization-using-r-and-data-extracting-from-sql-server/
2020-09-26 05:59:43 <dsal> I've never used R with a database.
2020-09-26 05:59:50 × thir quits (~thir@p200300f27f0fc60094e773283d7bf825.dip0.t-ipconnect.de) (Ping timeout: 246 seconds)
2020-09-26 06:00:01 <Guest26> but if i dont use R, but a haskell powered webpage instead, maybe i dont need SQL
2020-09-26 06:00:30 → mu_ joins (~mu@unaffiliated/mu)
2020-09-26 06:00:39 <fraktor> What is this data? I'm really trying to understand what kind of data would have the properties you're describing.
2020-09-26 06:00:41 <Guest26> the most important feature is to be able to somehow have the webpage read some data written to disk by the other haskell program producing the data
2020-09-26 06:00:44 <dsal> You're just using words in places where they don't fit and that's a bit confusing.
2020-09-26 06:01:07 <dsal> Web servers are pretty good at reading data from disk and serving it to web browsers.
2020-09-26 06:01:21 <Guest26> % utc
2020-09-26 06:01:21 <yahb> Guest26: ; <interactive>:111:1: error: Variable not in scope: utc
2020-09-26 06:01:46 <Guest26> % getCurrentTime
2020-09-26 06:01:47 <yahb> Guest26: ; <interactive>:112:1: error:; * Variable not in scope: getCurrentTime; * Perhaps you meant `getCurrentCCS#' (imported from GHC.Prim)
2020-09-26 06:01:58 <Guest26> hmm, thats from Data.Time
2020-09-26 06:02:29 <Guest26> [(UTCTime,Double)]
2020-09-26 06:02:52 × Saukk quits (~Saukk@2001:998:dc:4a67:1c59:9bb5:b94c:4) (Remote host closed the connection)
2020-09-26 06:02:57 <fraktor> I understand that's the format of the data.
2020-09-26 06:03:02 <dsal> % getCurrentTime
2020-09-26 06:03:03 <yahb> dsal: 2020-09-26 06:03:02.989582121 UTC
2020-09-26 06:03:06 <fraktor> But like, what are you trying to accomplish with this software?
2020-09-26 06:03:07 → berberman joins (~berberman@2408:8207:256c:4a80:584e:a9ff:fe9b:d3fe)
2020-09-26 06:03:21 <Guest26> maybe i could write each data entry to its own file, but that seems bad. thats where i thought an SQL database that could be simultaniously updated by one process, and read by another process, would be good
2020-09-26 06:03:22 <fraktor> I'm asking because a solution might already exist for what you're doing.
2020-09-26 06:03:40 → heatsink joins (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net)
2020-09-26 06:03:53 × ahri quits (~ahri@178.209.40.84) (Quit: Ping timeout (120 seconds))
2020-09-26 06:03:58 <Guest26> certainly not, the data is being calculated by another program, its impossible this would already exist
2020-09-26 06:04:05 <dsal> Do you have a problem you're trying to solve? It just sounds like you have a few parts y ou want to be part of a solution, but you don't have a problem that will use all those parts.
2020-09-26 06:04:14 <Guest26> you could think of it as a mass spring system for example
2020-09-26 06:04:23 <fraktor> Oh, so you're running simulations?
2020-09-26 06:04:26 → ahri joins (~ahri@178.209.40.84)
2020-09-26 06:04:39 <Guest26> its just calculated data, that is output infrequently
2020-09-26 06:04:59 × adamwespiser quits (~adam_wesp@209.6.42.110) (Ping timeout: 240 seconds)
2020-09-26 06:05:00 <Guest26> and i want to have the time on the x axis of the plot in the html page
2020-09-26 06:05:13 <fraktor> I see. So you do want a time-series database, but Prometheus is kinda the wrong tool.
2020-09-26 06:05:22 <Guest26> so that the most up to data version is displayed on refreshing the page
2020-09-26 06:05:59 <dsal> If you serve a file from a web server, it will serve the latest version of that file when requested.
2020-09-26 06:06:09 <fraktor> May I suggest Grafana? It's an open-source visualization tool that can use whatever as a backend, I think including static files
2020-09-26 06:06:10 → o1lo01ol1o joins (~o1lo01ol1@bl8-213-81.dsl.telepac.pt)
2020-09-26 06:06:12 <Guest26> the components i need are something equivalent to an SQL database for storing the data to, and something like a HTML page, that can do the visualization
2020-09-26 06:06:27 <dsal> Yeah, I'm using grafana with influxdb. It's pretty great.
2020-09-26 06:06:39 → justanotheruser joins (~justanoth@unaffiliated/justanotheruser)
2020-09-26 06:06:42 <dsal> I liked not having to write that part. heh
2020-09-26 06:07:03 <fraktor> In case you can't tell I've done DevOps stuff before, so Prometheus and Grafana are my go-to tools for system monitoring lol
2020-09-26 06:07:09 <dsal> I've got a bunch of bits of haskell that read and write influxdb and a bunch of that stuff ends up in grafana.
2020-09-26 06:07:32 <dsal> I've managed to never use prometheus. At this point, I'll just keep not using it .heh
2020-09-26 06:07:39 <Guest26> this? http://hackage.haskell.org/package/grafana-0.2 https://en.wikipedia.org/wiki/Grafana
2020-09-26 06:07:42 <Guest26> looks good
2020-09-26 06:07:53 × heatsink quits (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net) (Ping timeout: 246 seconds)
2020-09-26 06:08:25 <dsal> It's very unlikely you want a grafana haskell lib.
2020-09-26 06:08:29 <fraktor> I'd recommend using an intermediary DB now that I understand a little better, since that means you can run arbitrary queries on it.
2020-09-26 06:08:37 <Guest26> and this https://hackage.haskell.org/package/influxdb https://en.wikipedia.org/wiki/InfluxDB
2020-09-26 06:08:57 <fraktor> The Grafana haskell library is for when you want to write a custom interface with grafana, and using an intermediary database is simpler and more powerful.
2020-09-26 06:09:01 <Guest26> ah, yeah, i thought SQL seemed better than trying to write to the disk
2020-09-26 06:09:09 <Guest26> so is that where influxdb is advised?
2020-09-26 06:09:09 <fraktor> Yep. InfluxDB and Grafana are a good pair from what I've heard
2020-09-26 06:09:23 <Guest26> dsal: any chance of a demo?
2020-09-26 06:09:45 × mu_ quits (~mu@unaffiliated/mu) (Read error: Connection reset by peer)
2020-09-26 06:09:47 <dsal> "sql" and "write to the disk" aren't different things.
2020-09-26 06:09:51 <dsal> I use sql to write to the disk.
2020-09-26 06:09:58 <dsal> Let me see if I can find a good one.
2020-09-26 06:09:59 → mu_ joins (~mu@unaffiliated/mu)
2020-09-26 06:10:24 <Guest26> https://hackage.haskell.org/package/prometheus https://prometheus.io/
2020-09-26 06:10:30 <Guest26> im not sure about that...
2020-09-26 06:10:39 <Guest26> i dont think metrics is the right idea here
2020-09-26 06:10:51 <fraktor> Yeah, Prometheus is the wrong tool here
2020-09-26 06:10:53 <dsal> You've made it really hard to have any idea what you're doing.
2020-09-26 06:11:08 × o1lo01ol1o quits (~o1lo01ol1@bl8-213-81.dsl.telepac.pt) (Ping timeout: 272 seconds)
←Prev  Next→
Page 1 .. 217 218 219 220 221 222 223 224 225 226 227 .. 5022

All times are in UTC.