thanks so much guys, its up and running

info for all : i live in Egypt

@marcinozga , thanks for your suggestion , i'm starting to think about one big server instead of two.

@Pete-S , unfortunately, supermicro is not available here and the shipping cost for a new server is massive.

@dave_c , the cost to buy something from outside the country is tough, we cant do that right now.

@bnrstnr , the r510 is the only server i can find here with 3.5 slots, either that or older,
and we can't import from outside the country these days.

learning about Gluster and Ceph and which one is better for my needs to deploy

@DustinB3403 , is this question for me?
if it is, to me a proper storage server would be an enterprise server hardware with a stable server os, physical or virtual (i too prefer virtual), with a lot of storage features :D, and a scalable storage system for future expansion

thanks so much guys, its up and running

which confg file to edit and put the above configs??
and is these config if the nginx on the same server as RocketChat or on another server??

hi, thanks for the info, is there any new update on the matter??
can you share more info on how to setup rocket.chat and nginx on fedora 29

@DustinB3403 , is this question for me?
if it is, to me a proper storage server would be an enterprise server hardware with a stable server os, physical or virtual (i too prefer virtual), with a lot of storage features :D, and a scalable storage system for future expansion

@scottalanmiller said in Advice On a New Setup:

@GodfatherX64 said in Advice On a New Setup:

what setup do you suggest to create a failover backup cluster?
lets say i want to connect to the 2 on site servers with one connection or one path , an if one is down the second is up with the same data on it

If you want a failover cluster for a file server, use Starwind and do your failover at the platform level (hypervisor), not in the VM. For other workloads, like databases, you want the application to handle it.

@scottalanmiller , thanks

instead of the hypervizor failover, what do you suggest for a sync service in linux to just copy the differtintals between files, what do you think about rsync, would it be adequate than the hypervisor

info for all : i live in Egypt

@marcinozga , thanks for your suggestion , i'm starting to think about one big server instead of two.

@Pete-S , unfortunately, supermicro is not available here and the shipping cost for a new server is massive.

@dave_c , the cost to buy something from outside the country is tough, we cant do that right now.

@bnrstnr , the r510 is the only server i can find here with 3.5 slots, either that or older,
and we can't import from outside the country these days.

@scottalanmiller said in Advice On a New Setup:

@GodfatherX64 said in Advice On a New Setup:
Keep in mind even big shops when they talk about clustering for file servers it is for failover, not load balancing.

what setup do you suggest to create a failover backup cluster?
lets say i want to connect to the 2 on site servers with one connection or one path , an if one is down the second is up with the same data on it

wow, thanks guys for your informative opinions

so
@scottalanmiller , WD Gold, OBR10 for performance, fedora (vm) with xfs maybe,

@scottalanmiller @Obsolesce , ok , no RAID expansion, plan from the beginning , got it.

@marcinozga , unfortunately, like @scottalanmiller said it will be more costly than local, and the gigabit fiber is between 2 buildings , not the internet connection

@Pete-S , thanks for the 12tb suggestion, but i want performance too if i went for 10gb lan
and unfortunately, i couldn't find any 3.5 servers except these 2, r710, r510

Hi, i've been reading for last 3 weeks or so about sam-sd (open storage) but i can't decide which is best for my situation, and i would really appreciate your opinions on this setup as a (veeam backup, file archive and a lightly accessed file server) on site and an identical mirrored one off-site next building (1gb fiber)

Current Setup
2 old ibm x3650-7979 with raid 10 (6x 2tb wd black), please dont judge, it was already here when i started working here and we're moving forward :thumbs_up_medium-light_skin_tone:
but now it is so slow with backups and restores, so power hungry, raid is doing funny things and corruptions, low storage

Goal
Run a 'proper' storage server OS
Upgrade storage capacity
distributed file system between the two servers

thought of Dell R510 (Acctually, it is the only one beside r710 with 3.5 we can find as a refurbished and more than 6 hd bays)
single 6 core cpu
32g Ram
Raid H700
12 (empty bay)

What i could find in my county is these hard drives

Seagate IronWolf NAS 4tb, 6tb, 8tb, 10tb
Western Digital Gold 4tb, 6tb, 8tb, 10tb, 12tb
Western Digital Red Pro 8tb
Western Digital Red 4tb, 6tb, 8tb

what we can afford is 4tb, 6tb or 8tb of any model, unfortunately the increase in price between 8tb to 10tb or 12tb is out of our reach,

i would like your opinions on
1- Which first choice and second choice drive model to buy,
2- Which RAID setup if i want at least double the current setup capacity with the availability to expand the array later with just adding more drives and if it is OBR10, is it expandable on this setup or just the OBR6
3- Which linux (Virtualized) OS to use? im familiar with most of them
4- Which tool to use to mirror the two servers and distribute the load , can i use gluster or ceph or it is not worth it and just use DRDB

thanks in advanced guys.

learning about Gluster and Ceph and which one is better for my needs to deploy