vitessio / website Goto Github PK
View Code? Open in Web Editor NEWvitess.io website and other project management related content
Home Page: https://vitess.io
License: Apache License 2.0
vitess.io website and other project management related content
Home Page: https://vitess.io
License: Apache License 2.0
On page https://vitess.io/docs/overview/scalingwithvitess/
the following links don't work:
Horizontal sharding (Codelab)
Sharding in Kubernetes (Codelab)
They should probably point back to the tutorials - local and kubernetes respectively.
Fairly often I want to reference flags for vtgate, vttablet, etc. We already have decent descriptions for the help docs, so it shouldn't be too difficult to output those to the website. We could copy them out for now, but it's probably best long term to autogenerate them from the main repo so they don't go stale.
I think that this is the perfect example: https://kubernetes.github.io/ingress-nginx/user-guide/nginx-configuration/configmap/. It shows the flag, type and default value in an easy to reference table, hyperlinked to the description.
Instead of the blogspot
https://vitess.io/docs/reference/vitess-sequences/
We have the following content on this page which is incorrect. This needs to be corrected.
In a sharded keyspace, a Sequence’s data is only present in one shard (but its schema is in all the shards). We configure which shard has the data by using a keyspace_id for the sequence, and route all sequence traffic to the shard that hold that keyspace_id. That way we are completely compatible with any horizontal resharding.
Sequences are documented correctly in the tutorial:
https://vitess.io/docs/tutorials/kubernetes/
The sequence table is an unsharded single row table that Vitess can use to generate monotonically increasing ids.
Since sequences are unsharded tables, they will be stored in the commerce database.
Nothing urgent, but the third column navigation that exists in the user guides is missing from the vtctl and api reference pages.
https://vitess.io/docs/reference/vtctl/#executefetchasdba
The Vitess API reference doc looks old. It refers to v3 API as alpha. The entire doc should be reviewed and brought up to date.
We need to make it easy to archive user documentation like so:
/latest-version/sectionname
/2.2/sectionname
The edit link allows you to edit in GitHub, but I believe doing so will create commits without a DCO:
This appears to be a known problem:
todogroup/gh-issues#50
Maybe we should remove the edit link, and just have a create issue link?
Add a community page at vitess.io/community modeled after Envoy's.
Preliminary content in GDocs.
Introduction of third column has broken most presentation widths.
The "Edit" (in GitHub) links on the public docs are broken. For example, this page links to https://github.com/vitessio/website/tree/master/content/docs/_index.md but after #195 this should now be https://github.com/vitessio/website/blob/master/content/en/docs/_index.md.
cc @lucperkins
Social media in /docs page navbar throw out empty values, even though params are assigned correctly in the config file.
All links checked except for the links referring to sharding. Remember to fix.
Follow Jiten's list
Add a page under "Reference" which clearly explains MySQL compatibility story. Rename https://vitess.io/docs/reference/mysql-server-protocol/ to fit in this page.
GitHub's robots.txt explicitly permits indexing the branch called 'master'. This is fine for source code, but for docs it creates a duplicate content problem in search engines.
It is also better to have Search Engines index https://vitess.io/docs/ instead of GitHub because we can add redirection if content needs to be moved around.
I suggest we rename the default branch to 'prod' to fix this.
@lucperkins is this something you mind helping with?
See also:
https://stackoverflow.com/questions/15844905/how-to-stop-google-indexing-my-github-repository
new slides and info
It's not necessary, but it would be convenient if on the sharding page, there were a little calculator that would help you define your keyranges. I would input the number of desired shards, and it would output all of the shard names for an equal split.
We have created a Vitess Calendar that lists the Vitess Monthly Meetings and will list Vitess Meetups.
We would like to make this available on the Vitess website so that the Vitess Community can add it to their calendars and stay up-to-date on events.
To new logos found here
They need to be updated in:
Note the different styles (in white)
I noticed what seems to be a typo in the documentation for the "Run Vitess Locally" tutorial. It states the following:
if you ran make build while in $HOME/vt/src/vitess.io/vitess, then you should set: export VTROOT=$HOME/vt
Then later the same tutorial has us run a 101_initial_cluster.sh
script which in turn eventually runs a vttablet_up.sh
script. In the vttablet_up.sh
script we see references to VTROOT such as:
init_db_sql_file="$VTROOT/config/init_db.sql"
It seems based on this that VTROOT should be pointed directly at the vitess repo root folder, not to a folder that would precede src/vitess.io/vitess.
Please correct me if I'm wrong, but it seems like the documentation should instead say:
if you ran make build while in $HOME/vt/src/vitess.io/vitess, then you should set: export VTROOT=$HOME/vt/src/vitess.io/vitess
If this seems right I would be happy to issue a PR with the change to correct this.
Relating to the thread with @sabutai's question on Slack (referencing this issue):
sabutai [15 hours ago]
For others who may be interested: We've found a workaround for this: By using Eclipselink QuerySequence, we were able to define a query to Vitess Sequences tables directly. This not only prevented SELECT LAST_INSERT_ID() call, we were also able to reduce number of DB trips since the application could request a bunch of IDs from Vitess, let's say 1000, so it will make only one call per 1000 inserts.
action: put faq for common scenarios back online
In Run Vitess with Kubernetes
Many of the Kubernetes links no longer exist, but we may no longer need these instructions so I will leave it for now.
content/docs/doc
On the Vitess monthly call today, I proposed adjusting the Vitess documentation table of contents to use DITA information typing + common elements seen in other technical docs (slides).
This includes the following changes to structure:
Not in scope for now:
The Resources page has content up to October 2018. We should add the following recent content:
Please review "Run Vitess locally" -
Updates and insights from the Vitess team
Old site had GViewer, which is no longer working well.
I switched the presentations section to ViewerJS, with the files hosted in the static
directory.
The current dimensions / usage of ViewerJS is not working well on mobile devices.
@lucperkins can you help me look into optimizing the presentations page experience with ViewerJS or other tools?
https://vitess.io/docs/schema-management/vschema/#advanced-usage
Since we changed the local example (vitessio/vitess/examples/local), this part of the documentation is now out-of-date. The new local example does not actually illustrate the advanced usage, so we may need to enhance that.
Or possibly this will be covered by the vreplication example once we are able to mainstream that.
Admins seek ability to associate specific users with specific queries and overall usage.
Table ACLs enable this. And vitess exports per-user stats.
Vars are listed here: https://github.com/vitessio/vitess/blob/master/go/vt/vttablet/tabletserver/tabletenv/tabletenv.go#L35
Here's the config: https://github.com/vitessio/vitess/blob/master/go/vt/vttablet/endtoend/main_test.go#L174
More:
If you look in /debug/vars
you'll see a few variables that start with User
, like UserTableQueryCount
. It's a multi-dimentional export categorized by Table, User and query type. You'll find similar names exported as prometheus metrics.
If you see an incident, looking at these variables will help you quickly narrow down the root cause because of how fine-grained these stats are.
Once you've identified the table and query type, you can further drill down int /queryz
(human readable) or /debug/query_stats
(JSON), to see if it's a particular query.
In config.toml
Currently, vitess.slack.com in top right goes to a login page.
The proper link to joining Vitess Slack as a new user is here.
Or: bit.ly link.
Thoughts on how to do this? (Given that the slack link is coming from a param in the config file)
Via https://vitess.io/docs/schema-management/:
Note that this functionality is not recommended for long-running schema changes. In such cases, we recommend to do a schema swap instead.
The link to schema swap is also broken. We should change this to:
Note that this functionality is not recommended for long-running schema changes. In such cases, we recommend to use pt-online-schema-change or gh-ost instead.
Replace: "getting started with Kubernetes" with just one command.
@sougou - what is it? :)
Hi Team,
Please help me with the steps of configuring multiple vtgates in my local machine through docker.
Is it possible to do so?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.