Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rfurlan
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
61.
▲
Ask HN: Bootstrapped and profitable - what would you like to know?
23 points
by
rfurlan
15y ago
|
39 comments
62.
▲
by
rfurlan
15y ago
You can't gain confidence by studying, you must gain confidence by doing. My suggestion is that you start building things, pick an idea and execute it begin-to-end. Then pick a more complex project and do it all over again. It is very impor
63.
▲
by
rfurlan
15y ago
Yes, we are aware and we can't wait to upgrade, we will probably do it at our next scheduled downtime :)
64.
▲
by
rfurlan
15y ago
We have never used SSRS, instead we have been slowly building an in-house analytics suite that reports all the information we need in real-time. Our analytics package is also based on MongoDB and it currently aggregates over 60 million even
65.
▲
by
rfurlan
15y ago
Yes, these are very useful features!
66.
▲
by
rfurlan
15y ago
I am actually just finishing writing an article with all the details, I will post a link as soon as it is online :) We are not using any middleware and all our queries are hand-sculpted to make sure we are never causing more I/O impact than
67.
▲
by
rfurlan
15y ago
We are currently not paying for support, although 10Gen is a pleasure to work with even when you are not paying them anything - they are very helpful! Thanks for the tip, we might do that eventually :)
68.
▲
by
rfurlan
15y ago
What is really important is keeping your indexes in RAM, our data already greatly exceeds the amount of RAM we have available. Even our indexes are only partially in memory already and performance is still terrific.
69.
▲
by
rfurlan
15y ago
That is an excellent question. It is trivial to back up a live SQL Server instance because it supports shadow-volume snapshots. The same isn't true for MongoDB. Instead, we had to resort to running our MongoDB database on a 3-node cluster a
70.
▲
by
rfurlan
15y ago
Regarding requirements, we were looking for something that was free and open source because we are planning to scale up our operations in the near future. Because we have both Windows and Linux servers, we also wanted something that could r
71.
▲
by
rfurlan
15y ago
Yes, we did had to invest on new servers. Not only we need more RAM, we also needed more storage (which as expensive because we use SSDs). All things considered, SQL Server licences are not cheap either (6K-7K each) so in the long run I bel
72.
▲
by
rfurlan
15y ago
We always felt that using a relational database to power our service was a bit like trying to fit a square peg in a round hole. Every application is different but in our case, we would gladly trade off consistency for performance and schema
73.
▲
by
rfurlan
15y ago
Sorry, I forgot to answer that, we are running mongod :)
74.
▲
by
rfurlan
15y ago
Just to clarify it: MongoDB main target platform is Linux, the Windows version is clearly a second-class citizen at this time. Not only the Windows version performs poorly under I/O pressure, it also crashes and leaves the database in a cor
75.
▲
by
rfurlan
15y ago
Yes, we evaluated several other alternatives. Ultimately, we felt that MongoDB's was at the sweet spot of best fit (to our needs) and maturity. We tried to run Mongo on Windows and that was a bit of a disaster so we are running it on Linux.
76.
▲
by
rfurlan
15y ago
Books are great to get you started, but real-world problems will always be your best tutors. Try to build something by yourself begin-to-end, Google whenever you get stuck, rinse repeat for a while and before you know it you will start to f
77.
▲
by
rfurlan
15y ago
We practically rewrote the whole application. Some code was spared, but I would say that less than 20% of the original source was reused. This was not just because of the migration to MongoDB but more because we decided to take the product
78.
▲
by
rfurlan
15y ago
Overall, it went well, bumpy at first because we had to learn about Mongo's performance characteristics. Having a full-size database to work with made it a bit easier to identify friction points early on. To make the most of it, we bit the
79.
▲
by
rfurlan
15y ago
We just migrated a large production database for an existing website to MongoDB. I am currently writing an article to share the lessons learned, in the mean time I would be happy to answer your questions!
80.
▲
Ask HN: We just migrated from SQL Server to MongoDB - ask us anything
26 points
by
rfurlan
15y ago
|
32 comments