08:12:56 GMT <picatta> Hi, just a question here: Is it possible to re-route an existing asp.net application to connect to Redis thinking that Redis is a MSSQL instance? So that Redis can cache the results from MSSQL?
08:30:27 GMT <Inge-> no
08:30:36 GMT <Inge-> and their structures are very different
08:53:44 GMT <picatta> --> Inge, so its impossible hmm... Do you know about any other methods that I could look into?
09:16:30 GMT <Inge-> What problem are you trying to solve?
09:33:08 GMT <picatta> I have an external asp.net application that asks for the same kind of data, this data takes a lot of time to generate. I have no access to change the app or sql queries, but if I could cache those datas I think that we could get a huge performance boost.
09:43:14 GMT <Inge-> well, it would seem to be asking for the data in its current relational-database form
09:44:03 GMT <Inge-> so not sure what you can do about the actual generation. You could possibly do some optimization of the MSSQL database - such as indexes, RAM/pining tables in memory and other things
09:44:29 GMT <Inge-> but you can't magically translate sql queries sent to an mssql database into redis queries in any way I know of :)
09:49:50 GMT <arjen-jonathan> Hi all
09:50:18 GMT <arjen-jonathan> I'm trying to figure out how to setup a priority Queue with FIFO behavior for values with the same prio
09:50:30 GMT <arjen-jonathan> Any good ideas?
10:04:11 GMT <Inge-> how about one queue per priority level, and always check higher priority queues before lower priority queues?
10:06:27 GMT <arjen-jonathan> Inge-: I thought about that, but the problem is that I'd like to use a blocking pop to watch the queue
10:06:54 GMT <arjen-jonathan> And that doesn't really work with that setup
10:15:40 GMT <minus> write a redis module to do priority queues :D
10:16:05 GMT <minus> i vaguely remember something about multi-block support in there. i might be wrong though
10:20:24 GMT <arjen-jonathan> In where?
10:20:30 GMT <arjen-jonathan> What would multi-block do?
10:20:52 GMT <arjen-jonathan> You can watch multiple keys, but there is no order among them
10:28:01 GMT <minus> ¯\_(ツ)_/¯
10:28:12 GMT <minus> probably could do something with a sorted set and keyspace notifications
10:29:20 GMT <minus> people must've done priority queues in redis before
10:58:00 GMT <picatta> thanks for the reply Inge--
11:05:14 GMT <Habbie> arjen-jonathan, how about a BLPOP on a semaphore key that just tells you "there is new stuff"?
12:51:51 GMT <wili_> Hi, is it possible to use HINCRBY, that command wont do anything if entry is not exists?
12:52:02 GMT <wili_> or do I have to use exists before?
14:05:33 GMT <Habbie> wili_, i see no flag on it to only incr if it exists
14:05:45 GMT <Habbie> wili_, so you'll have to check, in a transaction perhaps, or push some lua
14:06:34 GMT <wili_> yes
14:07:19 GMT <wili_> btw if I want to use redis with limited mem size, https://redis.io/topics/lru-cache is my way?
14:07:36 GMT <wili_> like if size will be larget data would be save on hdd
14:07:40 GMT <wili_> (building crawler)
14:10:22 GMT <Habbie> redis has no 'disk overflow'
14:10:22 GMT <Habbie> if you hit the limit you set, anything above it just goes away
14:10:42 GMT <Habbie> if you want 'disk overflow', pick another database
15:03:30 GMT <ash_mobile> Does anyone use redis with an RDBMS for anything other than caching?
15:05:48 GMT <King_DuckZ> ash_mobile: I use it as a regular storage backend in my project here https://github.com/KingDuckZ/dindexer
15:07:06 GMT <ash_mobile> I am just wondering if I already have an RDBMS if it makes sense to incorporate Redus into my stat
15:07:26 GMT <ash_mobile> Stack*
15:07:41 GMT <ash_mobile> Redis
15:08:51 GMT <ash_mobile> King_DuckZ ^
15:12:50 GMT <ash_mobile> So far I've only read articles that describe caching the latest results (and then filtering that information)
15:33:39 GMT <King_DuckZ> idk, in my case I use it as an optional backend
15:34:38 GMT <King_DuckZ> I don't know how useful it would be alongside another db...
16:56:17 GMT <hermatize> anyone familiar with splitting up tasks for workers with python rq
17:10:44 GMT <hermatize>  anyone familiar with splitting up tasks for workers with python rq?
17:11:08 GMT <minus> we heard you the first time
17:11:15 GMT <hermatize> sorry
17:33:39 GMT <hermatize> anyone even in here?
17:33:54 GMT <hermatize> need help with parallel tasks
18:07:46 GMT <wili> hi If I want to perform delete and hmset do I have to call them i transaction? There is condition that I want to perform hmset only when delete succeed
18:08:36 GMT <wili> (I am using redis-py)
18:10:04 GMT <minus> best course of action is to use a lua script
18:10:28 GMT <minus> otherwise a WATCH/MULTI/EXEC triple, but that can fail
18:11:13 GMT <wili> pff lua script, i should learn some :D
19:58:34 GMT <hermatize> whats this channel for really?
20:02:24 GMT <hue> hermatize: this channel is for redis
20:13:48 GMT <minus> really?
20:15:08 GMT <minus> the problem with your question is that it's not directly redis-related, hermatize. that dramatically decreases the chance that anyone here (who's not totally inactive) has dealt with it
20:42:28 GMT <snappy> q: does redis have some kind of trie-like data structure: patricia/radix/critbit trie maybe?
20:42:30 GMT <snappy> tree
20:47:38 GMT <snappy> hm, doesnt look like it, the preferred way to deal with prefixes is a sorted set, oh well
20:49:16 GMT <minus> snappy: Redis 4.0 coming; implement your own as module :D
20:50:16 GMT <snappy> minus: tempted :)
20:50:32 GMT <snappy> actually very much so
20:50:46 GMT <minus> no idea how feasible it is, but interesting for sure
20:51:40 GMT <snappy> well there's some great code for crit-bit trees out there, but they do some hacks for efficiency (reduce size of nodes by stashing data in aligned pointers etc.)
20:51:59 GMT <snappy> does redis 4 have some kind of module system?
20:52:22 GMT <minus> yup
20:52:45 GMT <minus> http://www.antirez.com/news/110 https://github.com/antirez/redis/blob/unstable/src/modules/INTRO.md
20:54:00 GMT <snappy> that's pretty cool, i'll epxeriment with it tonight, thanks for the heads up
22:30:52 GMT <sphinxo> is there an RedisModule_OnExit function, to do cleanup? ( when writing redis modules )
22:33:15 GMT <Habbie> on exit of what? redis?
22:43:03 GMT <sphinxo> Habbie: yes
22:43:15 GMT <sphinxo> Habbie: actually nvm, it's not an issue for me
22:43:26 GMT <Habbie> sphinxo, i have no idea but there's always atexit()
22:43:41 GMT <Habbie> sphinxo, don't know if that's safe in redis though