2830
English speaking PostgreSQL public chat. This group is for discussions on PostgreSQL-related topics and strives to provide best-effort support as well.
To demand from self is a really tough task, I know.
Читать полностью…
Perhaps maybe consider bunny.net as they provide connectivity and actually started as providers of other art archives :P https://bunny.net/pricing/cdn/#calculator
Читать полностью…
e.g. west coast South America might actually be better served by California and Dallas than Sao Paulo due to the dearth of interconnections in the center of the continent - they get their connections via submarine cable instead.
Читать полностью…
Yes, I was considering that, it seems the most obvious solution aside from a surrogate key (related to the ::text cast mentioned in the first thread - actually could probably just define it as an expression index, but not sure that would fit the logical replication requirements).
Читать полностью…
I'm a bit clunky and never used logical replication (i am a big fan of normal log shipping + replication slot), but can't you add a GENERATED AS column, cast the box to something that has equality (e.g. TEXT) and use that as the key? It is not elegant at all, but i assume it would stay the same at least between main pg versions
Do you actually run inkbunny, or are you running a mirror of inkbunny?
Читать полностью…
You must have a primary key on a table.
If the data isn't suitable as keys — use surrogate keys.
(Ask you db architect for a proper relational decomposition of your database. 3rd nornal form is a bare minimum to begin with.)
Immediate Joiner – Power BI Developer | 5+ Years ExperienceH
ello, I am Mohammad Asgar, an experienced Power BI Developer / Data Analyst with over 5 years of expertise in:
Power BI (Advanced Dashboards, Drill-Down, Drill-Through, Storytelling)
SQL, DAX & ETL
Data Modeling
Delivering KPI-driven reports across Telecom, Retail & Manufacturing domains
Delivered enterprise-level dashboards improving decision-making by 10–20%
Proven success in handling large-scale datasets & cloud migrations
Immediate Joiner (within 3–4 days) | Open to Relocate PAN India
LinkedIn: https://www.linkedin.com/in/md-asgar-95425162
9902232976 | asgariitb@gmail.com
Actually, the documentation of PostgreSQL is quite good. I would start there with the topics that are most relevant to your current project.
https://www.postgresql.org/docs/
Aurora postgresql is a database engine made by Amazon that has a "postgresql compatible" interface
Читать полностью…
I chose this method because we just need a best guess and we can IP geolocate, most are "reasonable" choices if not always optimal. And it fits nicely into PostgreSQL's box matching.
Perhaps I can 'just' create a hash index for box with something like create operator family 'box_hash_ops' using hash and create operator class for data type 'box' using 'box_hash_ops', which only has one required and two optional support functions (basically need to compute a stable hash for 32 and ideally 64-bit).
You'd need to either have all of your servers traceroute to your user, or keep a table of subnets and AS peerings and "guess" the route, or have your client do it client side (which sounds like the most sensible way honestly)
Читать полностью…
We don't know how many hops it is to an arbitrary viewer, and calculating the latency would be interesting... I do manually estimate latency when setting the areas using mtr.
Читать полностью…
But that also makes me think: is that part of an algorithm to decide from where to serve media? I would move it to a hop based or latency based one, unless there's legal reasons (e.g. geofencing) to do so
Читать полностью…
I really run Inkbunny. And we have always run on PostgreSQL, which is something I have been very thankful for on numerous occasions, despite occasional setbacks. 😼
Читать полностью…
I run a furry art archive; I am the DB architect. Well, since the founder handed it to me a decade ago. We don't really have funding for specialists, just $10/day for hosting. 😼
I guess that works, and thank you for suggesting it, just annoyed because the data is suitable, it just lacks the equality opclass and so I'm wondering if it is definable via an extension etc. (Perhaps others ran into this? Though I figure box isn't hugely popular. Ah, more from a decade ago... and the type is ~40 years old, the original Postgres.)
I just ran into the issue that box doesn't have an equality operator, so can't be updated in a logical replication setup (it breaks replication until you skip the transaction if you do update a row on the primary).
We use them (logically) for catchment areas for a list of cache servers. Is there any good way around this other than removing the table from the subscription. Can we define our own box equality opclass that is equality of its components rather than equality of areas, say? (Because a box in Africa is not the same sized box in America.)
(Just to make it more difficult, although perhaps not for this purpose, multiple caches can cover the same areas, usually at different priorities; the primary key if any is a composite of cache ID and box. Different areas for the same cache may also have the same priority, although we could potentially forbid that and run an index on the composite of cache ID and priority - not sure that works though, and it's a dubious solution.)
The postgres has good docs... And I would suggest to read them through carefully, at least once, to everyone willing to be a postgres DBA or postgres user...
But that wouldn't be sufficient to be a good DBA. Some experience with less responsibility seems to be required.
(As for MS SQL, but there there are more training courses to begin with).
Another problem is tgat Aurora isn't postgres. While Amazon RDS for Postgres is mostly a closed fork of Postgres, with some rare glitches — the Amazon Aurora definitely isn't.
If you'd be lucky — the AWS team would do most of the work for you.
If you'd not — not a much you could do...
Even though I agree it is not OSS Postgres, it’s not just a Postgres compatible interface service like DSQL or Spanner. Linking this initiative from @mayur555b https://pgscorecard.com/
So ChillBRO I would recommend to first search for ressources on how to master Postgres as a SQL Server DBA and then look at the specificities of Aurora. There are plenty of good stuff out there
But the migration team is saying that Postgres aurora
Читать полностью…
And in our environment we are using SQL Server replication among servers as it’s our business need
Читать полностью…
All…we are in planning to migrate/move from MS Sql to Aurora PostgreSQL, what should I learn and where to start
Читать полностью…
also, next time forbid auto DDL, plan for a proper source control and deployment procedure for DDL scripts
Читать полностью…
