planned outage Jan 11-12
 

Hey people.... If all goes well, I will be doing some more improvements on the infrastructure stack for the weekend of January 11-12. So if the network is down, check Facebook for status posts. I finally got tired of backups stalling and deleting VMware snapshots taking hours. So the storage network is being upgraded from 1 gig to 10 gig.

Re: planned outage Jan 11-12
 

Special Chris,

Your ongoing diligent efforts of such long standing deserve a thank you from all the participants of this forum. Unfortunately there will be many who who do not know pf the length and depth of your involvement and generosity.

I for one am absolutely amazed that yo have personally kept the site running for so long and at your own expense. There can not be anything comparable within cyberspace. Money could not have bought what you have provided. We can only very much hope, that you have gained huge satisfaction as a result of your selfless hard work.

You more than deserve a happy and prosperous New Year. I will be there before you so expect is a sincere welcome. LOL

Trevor.

Re: planned outage Jan 11-12
 

Thank you for the continued hard work and lets hope for a smooth transition.

Was the snapshot consolidation due to network latency? Wouldn't that be related to disk performance instead? This is without knowing how long/old they are but just a question.

Re: planned outage Jan 11-12
 

I'd like to offer my thanks as well. Seeing other forums being bought up by that crap company verticlescope is disheartening. I'm really glad this site is still unchanged and going strong. Thank you Chris!

Re: planned outage Jan 11-12
 

I too would like to thank Chris for his efforts. I am relatively new in these parts, but I certainly would never have bought my SVX without knowing that this resource was available. My thanks also to the many members past and present whom have contributed to this valuable resource.
Happy new year to all!

Re: planned outage Jan 11-12
 

Thanks Chris.

Re: planned outage Jan 11-12
 

In short, the SVXWN runs as a VM on a Cisco UCS-C220m4 server. It does not use local disk, but the storage LUNs are presented as block storage via iSCSI from a QNAP SAN over a network to the Cisco server. The network speed is 1gbps but the theoretical storage speed limit in megabytes is 125MBps, and we average 117MBps. The issue is snapshot management and weekly backups as this limit is rather slow vs local disk that would average 300-500MBps or more. So, the network transport that is being used at 1gbps i being upgraded to 10gbps to eliminate the bottleneck. As users, you will likely never see the the bottleneck, but it would be nice to make sure it never becomes an issue. What I am trying to do is make backups take less time because spooling to tape becomes an issue if the data hitches. Also moving large packets of data around in VMs gets troublesome at the slower speed. Hope this helps.

Re: planned outage Jan 11-12
 

Good info for those curious about the infrastructure. Is the storage shared or dedicated?

Re: planned outage Jan 11-12
 

Image of the infrastructure running the SVXWN :D

http://www.subaru-svx.net/photos/files/Chris/44915.

Tour for anyone that is interested.... or bored
Top down:

• Cisco ASA5520 (Firewall and VPN)
• (space - air for rear mounted Cisco 3750G gigabit switch (some storage traffic, mostly IP traffic)
• Cisco Nexus 5010 Datacenter Enterprise 10gigabit switch (Work in progress for iSCSI storage SAN traffic and fabric switching)
• (space - spare cables, power supply, wireless keyboard, etc...)
• Cisco UCS-C220m3 Enterprise class host (VMware 02)
• Cisco UCS-C220m3 Enterprise class host (VMware 01)
• (work in progress, these are the new homes to the SVX World Network and when all planned work is completed, there will be HA or high availability so that each of the machines backs the other up for failures)
• HP Proliant DL360 G7 (This machine is being used to run the SVXWN VMs non-disruptively while the work is being completed on the Cisco C220s. When completed, this machine will be connected to a tape library for redundant backup jobs for almost 20 years of data)
• (space - spare hard drives and SSDs for QNAP.... got to have spares)
• QNAP Expansion shelf (48 terabytes of added capacity for VMware snaphots and data backups)
• QNAP Controller SAN (24 terabytes of hybrid SSD/spinning disk network storage array.... This feeds all the machines block storage)
• Dell Poweredge 350 (server running software routing, load balancing, and access lists both white and blacklists)
• Dell Poweredge R610 (meh... personal media PLEX server and game server... dont judge me!)
• Old as dirt Dell Poweredge MC1655 blade server (was the home of the SVXWN from 2006 until 2019. Robust and never let me/us down. One last blade to be migrated off of then full decommission to make room for Quantum Superloader 3 tape library)

Questions welcomed

Re: planned outage Jan 11-12
 

What's that in English? ;)

But seriously folks! I work in a variety of locations, all of which have rooms full of stuff like that. Some of the rooms are very large, have multiple floors, and signs for people who might be crawling around in them. I visit them but tend not to spend much time there; my work is mostly at the keyboard.