Silk Road forums
Support => Feature requests => Topic started by: DrDeepWood on November 26, 2012, 07:48 am
-
Why not seperate the vendor and user accounts on different servers? Vendors probably cause like 1% of the traffic and have a lot more to lose than users, also this will prevent orders from not going through in a situation like last week.
-
A vendor without buyers is a hermit IMO.
I can't see the advantage for separation of buyers and vendor accounts on different servers as what really matters is the GUI and the database behind it containing all the orderinfo etc. When this engine is down, the road is blocked;)
-
i think switching the httpd would improve alot.. apache ulimit is craphola, go for nginx and set up a reverse proxy to filter traffic through a IDS (snort_inline for example) to filter malicious attacks.
just an idea
-
A vendor without buyers is a hermit IMO.
I can't see the advantage for separation of buyers and vendor accounts on different servers as what really matters is the GUI and the database behind it containing all the orderinfo etc. When this engine is down, the road is blocked;)
a redunant sql cluster would be good indeed. I am afraid they don't use remote sql clusters and all databases are stored locally.
offtopic: another suggestion is the ability to download the wallet.dat in case SR gets raided or anything in that way.
-
I would love to see vendors and buyers being on separate servers if that is possible, maybe even a third server to actually hold the physical orders (meaning have the buyer server connect to the order server and have the vendor server query the order server when needed). that would reduce a lot of traffic and might even avoid some down times.
-
I would love to see vendors and buyers being on separate servers if that is possible, maybe even a third server to actually hold the physical orders (meaning have the buyer server connect to the order server and have the vendor server query the order server when needed). that would reduce a lot of traffic and might even avoid some down times.
remote sql clusters and a few webservers running nginx, not shitty apache