Looker Connect Training

Help Center

Documentation

Community

Cloud Certifications

The Kitchen Table

Explore the product knowledge we've built up together

2110 Topics

Taylor_O_Meara
Looker Staff
Taylor_O_MearaLooker Staff
published in Technical Tips & Tricks

How to overwrite content with GazerKnowledge Drop

Last tested: Mar 25, 2019 Gazer is a command line tool for Looker content management. You can use it to overwrite existing content with new content so that the URL of the existing content does not change with the update.How to do itSet up:Old dashboard: ID=15, name=Old Dashboard New dashboard: ID=19, name=Renovated Dashboard instance = example.looker.comRequirements:Both dashboards in the same spaceStep by stepCopy the content of the new dashboard into your current directory. gzr dashboard cat 19 --host=example.looker.com --dir ~ Then you can ls to see the new file: it should have the same name as it does in the instance (that's what "~" does), prepended with "Dashboard_[ID]_". Note that this example "Renovated Dashboard" has a space, which the terminal will store as "\ " in subsequent commands. Open the new dashboard file in any text editor. vi Dashboard_19_Renovated\ Dashboard.json Make 2 changes to the dashboard in the text editor. Remove the slug. It will look something like thi

4910
Taylor_O_Meara
Looker Staff
1 year ago
thomas.banghart
Looker Staff
thomas.banghartLooker Staff
published in Technical Tips & Tricks

Why are marketplace/custom visualizations throwing a `document is sandboxed and lacks the 'allow-same-origin'` error?Knowledge Drop

Last tested: Mar 24, 2021 You might see the following with an error similar to DOMException: Failed to read the `cookie` property from 'Document': The document is sandboxed and lacks the 'allow-same-origin' flag when viewing a marketplace/custom visualization after updating to 7.20:​Chances are you had the "Sandboxed Custom Visualizations" labs feature enabled prior to the update and will need to disable the newly introduced in "Unsafe Custom Visualizations" legacy feature: This can also come up with a message like Uncaught DOMException: Failed to read the serviceWorker property from Navigator: Service worker is disabled because the context is sandboxed and lacks allow-same-origin flag  This content is subject to limited support.                   

1110
thomas.banghart
Looker Staff
1 year ago
Vincent
Looker Staff
VincentLooker Staff
published in Technical Tips & Tricks

When looker is running in a clustered mode, where do we store the query cache?Knowledge Drop

Last tested: Aug 14, 2017 It’s on the local file system of whatever node ran the query, but node A can request it from node B.So if the user hits node A, node A looks up where the cache is (from the internal database), sees node B has it, requests it, node B sends it, and node A sends it to the user. This content is subject to limited support.                  

170
Vincent
Looker Staff
1 year ago
Vincent
Looker Staff
VincentLooker Staff
published in Technical Tips & Tricks

How many threads does the regenerator process have by default? Rather how many PDTs can it build at one time?Knowledge Drop

Last tested: Apr 23, 2021 This is controlled by the Max PDT Builder Connections setting in your database connection.Docs This content is subject to limited support.                   

120
Vincent
Looker Staff
1 year ago
Vincent
Looker Staff
VincentLooker Staff
published in Technical Tips & Tricks

max_cache_age and PDTsKnowledge Drop

Last tested: Nov 29, 2020 Question:If a PDT is tied to its datagroup that’s based on ETL with a max_cache_age of 24 hrs, if the ETL table doesn’t update, should that PDT rebuild after 24 hrs? Answer:No, max cache age has no effect on PDTs triggering. It only applies to cached explore/look/dashboard query results. And even then, it just drops cached results after 24 hours. It doesn't fetch new data.If you wanted a PDT to rebuild every 24 hours, use a sql_trigger_value that selects the current day. This content is subject to limited support.                  

3130
Vincent
Looker Staff
1 year ago
Vincent
Looker Staff
VincentLooker Staff
published in Technical Tips & Tricks

In clustered mode is it possible to have separate log files per node? or do they all write to the same file on the shared filesystem?Knowledge Drop

Last tested: Jan 21, 2019No, in a cluster environment, each node has their own log. This content is subject to limited support.                  

50
Vincent
Looker Staff
1 year ago
Vincent
Looker Staff
VincentLooker Staff
published in Technical Tips & Tricks

When a new version of a PDT is currently in the process of building, do we still use the old version?Knowledge Drop

Last tested: Dec 27, 2018 QuestionSay the regenerator runs the trigger query, and the trigger value of a PDT changes. Does the old table get dropped immediately or after the rebuild? What about when we change the code of a PDT, for example the sql_trigger_value , and push it to prod?AnswerIf just the trigger value has changed, and no code changes have been pushed to production, we will use the old version of the PDT until the new version is built. This will show up as "Using Old Version" in the PDT panel.If any change in the SQL code of the PDT has been pushed to prod, including the sql_trigger_value or datagroup code, we do NOT use the old version of the PDT—it is dropped immediately. Because the sql changed, the old version is no longer considered valid. The new one still has to build - that could mean a while before there is a valid PDT build in production.Good practiceIf you make a change to a PDT, build it in dev first, then when you push, that PDT will get promoted to prod. (Of c

510
Vincent
Looker Staff
1 year ago
Vincent
Looker Staff
VincentLooker Staff
published in Technical Tips & Tricks

Can I change the default location of the looker log?Knowledge Drop

Last tested: Sep 6, 2019 The only option is to symlink the log directory somewhere else.By default, Looker will save the logs to looker/log . The docs show this in the directory structure, and there's a feature request for allowing logs to be stored in BigQuery/Stackdriver. This content is subject to limited support.                  

70
Vincent
Looker Staff
1 year ago
William_Lane
Looker Staff
William_LaneLooker Staff
published in Technical Tips & Tricks

"Invalid Signature on SAML Response"Knowledge Drop

Last tested: May 3, 2018 This error points to an invalid certificate. The solution is to get the correct certificate from this section:<X509Certificate> ThisIsALotOfRandomCharacters </X509Certificate> And paste it into IdP Certificate in admin/saml like so:-----BEGIN CERTIFICATE-----ThisIsALotOfRandomCharacters -----END CERTIFICATE----- If this doesn't resolve it it's likely that the certificate is just invalid and that they need to figure that out. This content is subject to limited support.                   

1500
William_Lane
Looker Staff
1 year ago
V
Looker Staff
vishal.dharmadhikariLooker Staff
published in Technical Tips & Tricks

`Comparison method violates its general contract!` error when running a query.Knowledge Drop

Last tested: Oct 6, 2020 This is a pretty weird error that's popped up a couple times. This usually appears when we've pivoted data and there's a conflict in the inputs which causes a Java issue (Stack Overflow).One option to resolve this issue is to reduce the number of columns in the Explore. This content is subject to limited support.                  

1850
V
Looker Staff
1 year ago
V
Looker Staff
vishal.dharmadhikariLooker Staff
published in Technical Tips & Tricks

Can I clone a private git repo?Knowledge Drop

Last tested: Mar 4, 2021 Docs go over how to clone a public Git repository but can we do the same thing if our Git repo is private?Yes! We can do that following these steps:Create a new Blank Project instead of Clone. Click the 'Configure Git' button and use the existing private repo for the git remote. Continue with normal Project setup and follow the instructions, adding the deploy key to the repository. Create a new file in the empty project and create a commit (it can just be an empty file) Once the commit is made, you should see 'Pull and Merge other Changes.' Clicking on this will pull all the LookML from the repository into the project. You'll probably want to delete the 'dummy' file and at some point Deploy to Production. This content is subject to limited support.                 

2360
V
Looker Staff
1 year ago
V
Looker Staff
vishal.dharmadhikariLooker Staff
published in Technical Tips & Tricks

Are pbf tile formats supported in Looker when creating custom map layers?Knowledge Drop

Last tested: Nov 20, 2020Nope, we will actually need to use .mvt files. Please refer to this Community post on setting up the custom map layer. If you have .pbf files, you can use Tippecanoe to output mapbox vector tiles (.mvt) just as easily as it can do protobufs (.pbf). This content is subject to limited support.                   

450
V
Looker Staff
1 year ago
xin
Looker Staff
xinLooker Staff
published in Technical Tips & Tricks

Custom visualization error: Cross-Origin Read BlockingKnowledge Drop

Last tested: Mar 3, 2020 If you see "Cross-Origin Read Blocking" error for self-created custom visualizations, and the host url is https://raw.githubusercontent.com/....This is due to the fact that https://rawgit.com/ is no longer in service. You should migrate the js file into other services such as S3 bucket, GCS or https://raw.githack.com/ etc. This content is subject to limited support.                 

600
xin
Looker Staff
1 year ago
xin
Looker Staff
xinLooker Staff
published in Technical Tips & Tricks

Can I create a use a derived table in the subquery rather in the CTE?Knowledge Drop

Last tested: Jan 21, 2019 Yes you can. Actually this will not be a derived table any more, but the query will be faster. So how do I do that?The trick is with `sql_table_name`!Instead of using a table name in the `sql_table_name`, write the full subquery there, then define all the dimensions and measures as usual. The generated sql will be a subquery. This content is subject to limited support.                  

3770
xin
Looker Staff
1 year ago
xin
Looker Staff
xinLooker Staff
published in Technical Tips & Tricks

Can I convert table calculation from dimension type to measure type?Knowledge Drop

Last tested: Jan 21, 2019 Sometimes when we create a table calculation, we reference a dimension as condition, and in fact we want to use it as a measure type.The easiest way to do is referencing a measure value in the formula:count(dimension_type)+measure_type*0  This content is subject to limited support.                 

2490
xin
Looker Staff
1 year ago
xin
Looker Staff
xinLooker Staff
published in Technical Tips & Tricks

How to trigger datagroup/PDT rebuilds X minutes past each hour between certain hours of the day?Knowledge Drop

Last tested: Mar 5, 2021 The below examples are created for Bigquery.If you just want to trigger the datagroup/PDT rebuilds 10 minutes past each hour:if current minute is equal to or greater than 10, return the current hour, if not return current hour -1.SELECT IF(EXTRACT(MINUTE FROM CURRENT_DATETIME())>=10, EXTRACT(HOUR FROM CURRENT_DATETIME()), EXTRACT(HOUR FROM CURRENT_DATETIME())-1) 10 minutes past each hour between 9 and 19:Similar logic like above, however be aware that we use 18 rather 19, the last rebuild should be 18:10, and after 18:10, it should always return 18 until the next business hours starts at 9.SELECT IF(((EXTRACT(HOUR FROM CURRENT_DATETIME()) BETWEEN 9 AND 18) AND(EXTRACT(MINUTE FROM CURRENT_DATETIME())>=10)), EXTRACT(HOUR FROM CURRENT_DATETIME()), if((EXTRACT(HOUR FROM CURRENT_DATETIME()) BETWEEN 9 AND 18), EXTRACT(HOUR FROM CURRENT_DATETIME())-1,18 ))  This content is subject to limited support.                   

1090
xin
Looker Staff
1 year ago
zlebowitz
Looker Staff
zlebowitzLooker Staff
published in Technical Tips & Tricks

Why do we recommend many_to_one joins over one_to_many joins?Knowledge Drop

Last tested: Aug 2, 2019Minimize joinsBecause in most scenarios, the most granular table usually has the most valuable information, and is therefore the most likely one to always be necessary in your queries! For example, for a B2B SaaS company, most of your sales involve opportunity data (because that's where # of deals and $ ACV live). If you therefore start your explore with the Opportunity table, and then join into other tables (which usually have many_to_one relationships), your queries will on average contain fewer joins than if you started your explore with a less used table, like accounts.Avoid Symmetric Aggregates when possibleLooker is less likely to unnecessarily use symmetric aggregates if you define your explore from the most granular table. Let’s say you have order_items, orders, and users. You could define your explore starting with order_items (most granular) or users (least granular)Starting with the most granular table:explore: order_items { join: orders {   relations

9970
zlebowitz
Looker Staff
1 year ago
Zam
ZamMember
published in Technical Tips & Tricks

How to migrate SAML accounts when an email domain changesKnowledge Drop

Last tested: Jul 6, 2020 SAML has a predefined notion of a ‘nameid’ in the saml response document. We use that as the unique id for looker users. If you are not changing that per user, then all is good.If you are changing that, then you'll have to use the API to add credentials_email to all of the existing users that have the email of the new domain. Then when you move to the new domain, select the "merge users by email" option and it should be okay.Note: A colleague made an extension to help with this:https://github.com/davidtamaki/admin_power_packPlease note this is not officially supported resource - if you have a migration situation, please reach out by  filling an Issue in the git repo to get some help through it! This content is subject to limited support.                  

1560
Zam
1 year ago
xin
Looker Staff
xinLooker Staff
published in Technical Tips & Tricks

How to control the comparison arrow in a Single Value viz?Knowledge Drop

Last tested: Aug 30, 2018 According to the doc, positive numbers will display an upward arrow and negative numbers will show a down arrow. You can flip this so that "Positive values are bad".Beyond that, you'll want to use table calculations to manipulate your values accordingly.For some use cases, we might want to compare two values, for example: Sales value this month vs last month, we want to use arrow to indicate if this is an increase or not, we can use a table calc to do this using if() function.if(${last_month}>${this_month},${this_month}*(-1),abs(${last_month})) This content is subject to limited support.                  

1850
xin
Looker Staff
1 year ago
Department of Customer Love
Moderator
Department of Customer LoveModerator
published in Technical Tips & Tricks

How do Filter Suggestions Work?Knowledge Drop

Last tested: Aug 21, 2020 Filter suggestions are a way to save time when typing into filters and ensure you're choosing an option that exists in the data. When you click into a filter box, a list of suggestions will appear below the field. Where does this list of suggestions come from?Looker will run a query against the database to get all possible options for that field. It will look something like this:SELECT DISTINCT <field_name>FROM <table>WHERE (<field_name> LIKE '%' OR <field_name> LIKE '% %')GROUP BY 1ORDER BY 1LIMIT 1000 If the user types some characters into the filter box, Looker will substitute the appropriate conditions into the WHERE clause to filter the results. Looker will then show the first 100 of those results. Can I change what suggestions are populated?Yes! Check out this docs article. Are suggestions cached?Yes, they are! The cache length for filter suggestions is determined by the suggest_persist_for LookML parameter, which has a default va

7861
R
24 days ago
Department of Customer Love
Moderator
Department of Customer LoveModerator
published in Technical Tips & Tricks

Can you embed a Looker iframe in PowerPoint / Google Slides?Knowledge Drop

Last tested: Sep 8, 2020 Looking for a solution to export a dashboard of visualizations to a Google Slide deck on their template. Below were the options I was able to outline. Spoiler alert: none are ideal.(1) Python Script Overview Allows you to take a template slide deck and push viz images onto them Pros Most complete, detailed solution Works with Google Slides Very automatic solution once complete - would scale well Cons Requires significant dev time to create the solution Minor maintenance may be required to make changes over time (2) LiveSlides Overview Embed live visualizations into MS PPT presentation Pros Embed interactive visualizations into a PPT presentation Very easy to get set up Will update with live data with no work. No need to export images Cons Only available in MS PowerPoint Cannot be exported to PDF (3) Use a webhook tool like Zapier Overview Use a tool like Zapier to send a visualization from Looker and place it on slides Pr

19263
KarlR
1 month ago
Department of Customer Love
Moderator
Department of Customer LoveModerator
published in Technical Tips & Tricks

Why do large HyperSQL internal dbs cause problems?Knowledge Drop

Last tested: Feb 4, 2019 HyperSQL is an in-memory database. That basically means it needs to be loaded into memory to do its thing (store configuration, users, and other metadata). More specifically, the instructions to create the database (DDL, i.e.: data definition language) are persisted to disk. At startup, the script loads this into main memory and constructs the database in-memory. On a busy instance this database can grow in size, which leads to performance issues, Java memory pressure, and extremely long startup times. A concrete example is if more than the recommended 60% of main memory is allocated to the Looker java process, then the internal backend might be pinched to acquire the amount of main memory it needs to do its thing.https://docs.looker.com/setup-and-management/on-prem-mgmt/migrating-to-mysql This content is subject to limited support.                 

480
Department of Customer Love
Moderator
1 year ago
Department of Customer Love
Moderator
Department of Customer LoveModerator
published in Technical Tips & Tricks

What is cache_only_miss in the system activity History explore?Knowledge Drop

Last tested:  May 14, 2020 Looker has a feature that will fetch results from the cache if they existed before the page has fully loaded. If the cache only check does not find any results, we’ll mark that as a cache_only_miss.If we do find things in the cache there will be a complete record with a results_source of cache.If there is a miss (cache only miss) the queries will be executed again (without the cache only requirement) and the queries will be run from the database. This content is subject to limited support.                 

470
Department of Customer Love
Moderator
1 year ago
Department of Customer Love
Moderator
Department of Customer LoveModerator
published in Technical Tips & Tricks

Can I access, develop or extend the System Activity model?Knowledge Drop

Last tested: May 12, 2020 No, at this time users cannot access, develop or extend the system activity model.There is a feature request open for comment for this functionality. This content is subject to limited support.                 

511
IanT
1 year ago
Department of Customer Love
Moderator
Department of Customer LoveModerator
published in Technical Tips & Tricks

Is there a maximum value you can set user query limits to?Knowledge Drop

Last tested: Jan 21, 2019 No, there isn't a max per user query limit value.However, while there is no "max" value, as best practices we wouldn't suggest setting it higher than the connection pool size. Ideally, less than the connection pool size if you want to equally allow all users to run queries. This content is subject to limited support.                 

1440
Department of Customer Love
Moderator
1 year ago

Badges

  • Friendly Face
    sbrylowskihas earned the badge Friendly Face
  • Helpful
    carriehas earned the badge Helpful
  • Community Star
    carriehas earned the badge Community Star
  • Friendly Face
    carriehas earned the badge Friendly Face
  • Community Star
    Dawidhas earned the badge Community Star
Show all badges
Terms of Service