@goomba ehat problem specifically are you having?

:(

Quick question. When I am trying to restore a database from a remote dynamo table to a local dynamodb-local table, I keep getting these sorts of errors Caused by: java.util.concurrent.ExecutionException: clojure.lang.ExceptionInfo: :restore/read-failed Unable to read 58de91f6-6e59-4284-ad6b-94c262caf019 {:key "58de91f6-6e59-4284-ad6b-94c262caf019", :db/error :restore/read-failed}

2017-05-19 00:25:57.981 WARN  default    datomic.backup - {:message "error executing future", :pid 76, :tid 10}
java.util.concurrent.ExecutionException: clojure.lang.ExceptionInfo: :restore/read-failed Unable to read 58de91f6-6e59-4284-ad6b-94c262caf019 {:key "58de91f6-6e59-4284-ad6b-94c262caf019", :db/error :restore/read-failed}

etc

I just have no idea what this error actually means, so it's hard to troubleshoot, any insight would be greatly appreciated

@goomba does your MySQL give SELECT rights to user datomic and table datomic? Seems like it may not

Transactor needs select insert update delete

Peers just need select

It does, strangely

I can login to the datomic db with mysql -udatomic -pdatomic datomic and select * from datomic_kv just fine

😅

It's quite the mystery to me

okay what I've figured out so far is that the password parameter isn't being passed to MySQL for some reason

when I do tail -f /var/log/mysql/error.log I see the messages 2017-05-19T12:33:51.825580Z 316 [Note] Access denied for user 'datomic'@'localhost' (using password: NO)

@goomba What works for us here is adding &useLocalSessionState=true - i forget the where and why this was needed.

worth a shot!

eh... what still seems to be the case strangely is that it's only picking up the first parameter I pass in for some reason

it seems to ignore the rest of them

datomic: is what I just ran and it gives the same error 😅

If I switch the order of parameters around, it reads the password but not the username

Woohoo! Got it sorted! Need to escape that ampersand! Everyone can rest easy now 🙂 I may have gotten a little hint from cognitect sales team about being a shell scripting noob hehehe 😅

Has anyone tried to generate pull queries from s/keys specs?

also would like to throw in the question queue has anyone used the datomic console with a sql based storage?

struggling to figure out the correct command

datomic: is what I'm trying but no luck so far

you forgot ? @goomba

okay wasn't sure if that was important or not

yayyyy 😄 :thumbsup:

okay so if I make an entity of some sort and I decide I want it to have more attributes later, that's doable yes?

like this was presented in one of the examples

[{:db/ident        :story/title
   :db/valueType   :db.type/string
   :db/cardinality :db.cardinality/one}
  {:db/ident       :story/url
   :db/valueType   :db.type/string
   :db/cardinality :db.cardinality/one
   :db/unique      :db.unique/identity}
  {:db/ident       :story/slug
   :db/valueType   :db.type/string
   :db/cardinality :db.cardinality/one}]

these are not a single entity, these are three entities. This transaction creates the :story/title :story/url and :story/slug attributes @goomba

@goomba are you following along with a tutorial?

ah, yeah I'm trying to follow along with the "Day of Datomic 2016" videos

ok I see so my confusion was what an entity is

wait wait, just for clarification... aren't those three schema that collectively form an entity?

or do I have that backwards?

each schema attribute is itself also an entity

they don't collectively form anything

at the bottom, datomic is a set of assertions [entity attribute value transaction] (called "datoms")

an "entity" is just a number in the entity slot of a datom

oh fascinating

the map view of a datom is gathering together all datoms with a common entity

okay so if I wanted to (conceptually) add some information about a story, like an author, what would be the idiomatic way to do that?

query by some attribute of the story and then do some kind of association with the entity id and the new information?

by the way @favila I owe you at least one beer for all the help you've been! 🍻

first you need a reference to the story. This can be via direct entity id, or an attribute+value unique to the entity

e.g. [:db/ident :story/slug] or just :story/slug is the slug attribute entity.

so it looks like here :story/url is your unique attribute

so (d/transact conn [{:db/id [:story/url "the-url"] :story/author [:author/id #uuid"some-uuid"]}]) for example, would add a story author assertion to a story entity

the map form of the transaction is sugar for [:db/add [:story/url "the-url"] :story/author [:author/id #uuid"some-uuid"]]

so in one shot (d/transact conn [{:db/id [:story/url "the-url"] :story/author [:author/id #uuid"some-uuid"]}]) does a logical lookup on the story that matches "the-url" on the attribute :story/url and associates it with the :story/author specified... magical

now this presupposes that I've made a schema for :story/author first, yes?

yes

I was blind but now I see 😄 😄 😄

hi guys, im learning datomic using the official tutorial but unfortunately there is a query on the page http://docs.datomic.com/getting-started/see-historic-data.html which produces error can someone give me a hint what is wrong here? query: (<!! (client/transact conn {:tx-data [{:db/id commando-id :movie/genre "future governor"}]})) result: {:datomic.client-spi/request-id "b7ca85b3-b3d4-4f65-8f33-4b9d8b103fb5", :cognitect.anomalies/category :cognitect.anomalies/incorrect, :cognitect.anomalies/message ":db.error/invalid-lookup-ref Invalid list form: [[17592186045419]]", :dbs [{:database-id "591e005d-745c-492b-8f7b-ff406aaabda5", :t 1001, :next-t 1005, :history false}]}

@azimpel did you forget to ffirst the previous query result?

commando-id is [[123]] instead of 123

user=> (def commando-id 
            (ffirst (<!! (client/q conn {:query '[:find ?e 
                                                  :where [?e :movie/title "Commando"]] 
                                         :args [db]}))))
#user/commando-id
user=>

yes, you're the best

@favila , thank you, my dumb mistake

Is it guaranteed that the first datom in :tx-data off of a transaction report is the datom with the tx entity?

I doubt it

just look at the tx of any datom

they should all be the same for a given transaction

(-> tx-result :tx-data first :tx)

anyone knows what it could mean if (d/create-database db-uri) yields

IllegalArgumentExceptionInfo :db.error/read-transactor-location-failed Could not read transactor location from storage  datomic.error/arg (error.clj:57)

issuing (d/get-database-names db-uri) with a version of the db-uri that has a * in place of the datomic db name gives me nil which isn't unsurprising because when i look at the datomic_kvs table in postgres it is still empty

@tosh transactor is either not running or not pointed at storage correctly

or there are permission problems etc

Investigate transactor logs

ah of course, i haven't had a transactor running

thanks @favila worked like a charm. my first error was not starting the transactor (for some reason i thought the peer will write directly into storage and that starting the transactor only makes sense later once the datomic db is created) my second error was that in the properties file of the transactor I had the name of the datomic db that I wanted to work with instead of the name of the postgres db that i use for storage => so many db names and cryptic db urls 🙂

Is there a way to use transact-async without blocking on the future later?

You can block with a deref. Is your goal to call a callback when it's done (push not pull)?

Sorry should be "you can block with a timeout"

@favila yes I’d like it if it were fully async, i.e. callback based

Only client API has that. You would have to emulate on peer

is there a better/more-idiomatic way to do this?

(let [my-id (<!! (client/q conn {:query '[:find ?x
                                         :where
                                         [?x :story/title "Teach Yourself Programming in Ten Years"]
                                         ]
                                 :args  [db]}
                           ))]
  (println (<!! (client/transact conn {:tx-data
                                       [{:db/id              (ffirst my-id)
                                         :story/author-first "Peter"
                                         :story/author-last  "Norvig"}]}))))

(<!! (client/transact conn {:tx-data
                            [{:db/id              [:story/title "Teach Yourself Programming in Ten Years"]
                              :story/author-first "Peter"
                              :story/author-last  "Norvig"}]}))

@goomba unless you have a unique attribute (identity or value) you can include in the transaction map, the query for ID followed by transact is pretty standard

gotcha! Well I'm cool with that as long as I'm not writing code like a barbarian over here. Well, not a complete barbarian

It's often a good idea to have an externally unique identifier (ie domain ID of some kind) for this reason (and some other reasons)

In this case something like the library of Congress identifier or a UPC might be appropriate

Title prob not, since there are plenty of duplicate titles for different books

mhm. Unfortunately my weak human memory was having trouble recalling the UPC exactly so I tried to fudge with whatever I could remember 😂

Thanks again! 🎩's off!

Np

hey, how cool is our life that we get to do this? eh? eh?

still coding at almost 8pm on a Friday cause it's the 🐝's knees!

(if I was still stuck with JavaScript I would've hit the bar at 11am I think)

Lol

(11am pacific time, and I work on the east coast)

Is the backend of the site in datomic?

This is a roundabout way of asking if datomic is appropriate for serving videos

Oh wait just noticed they're using vimeo

But still, "Is datomic appropriate for serving videos?"

Definitely wouldn't store media in datomic

Not intended as a blob store

gotcha. not even close to my primary use case just curious

we have polyglot persistence for that sort of thing so not a big deal

Good for metadata for blobs stored in a secondary store like s3

yeah I was thinking something along those lines

blob store... I don't know where we come up with these words but another reason I love this industry 😂