Community
- Command Central
- MVP Program
- Monthly Mission
- Blogs
- Groups
- Events
- Media Vault
Products
- Observability
- Network Management
- Application Management
- IT Security
- IT Service Management
- System Management
- Database Management
Content Exchange
- SolarWinds Platform
- Server & Application Monitor
- Database Performance Analyzer
- Server Configuration Monitor
- Network Performance Monitor
- Network Configuration Manager
- SQL Sentry
- Web Help Desk
Free Tools & Trials
Store

Orion Resilience

After having our main Orion server die on us on Tuesday we are looking at various ways to make it more resilient.

My thoughts are hot standby pack for our 3 pollers and the database will now be on a SQL cluster my problem is what to do about the website.

Can the website go on a totally seperate box that hasn't got any pollers installed and if so how do I go about doing it, any other thoughts would be much appreciated

Thanks

Jon

IT Infrastructure Manager
Pilgrim Hospital
UK

Find more posts tagged with

Accepted answers

All comments

aLTeReGo

There is an "Additional Web Server" option for SolarWinds. Talk to sales. It lists for about $500.00 per web server.

m_roberts

The Hot Standby pack would provide the resilience for your Orion installation. If the primary Orion server goes down it will take over polling reception, feed data into database and provide the web interface. Once primary server recovers it will hand back.

Therefore if you are moving to a clustered SQL environment the only point of failure would be the additional pollers you have. In the main, with some manual intervention this is a relatively quick fix to move polling duties.

Hope this helps

------------------------------------------------------------
Prosperon Networks - Authorised UK SolarWinds Resellers
http://www.prosperon.co.uk

michael.buskin

Does the Orion Hot Standby have to run on another system and database though

-------------------------

If you have a problem click on email me with a brief description of the problem and / or solution

[url="mailto:michael.buskin@gmail.com?Subject=A Message from a SolarWinds Forum User"]Email Me

[url="www.v-eight.com"]Visit my Website

FormerMember

The hot-standby requires an additonal server but not another database. The way I understand how it works is...each poller continually updates, for lack of a better word, a "control" record in the SQL database (not sure of the sync interval but it's frequent). The "control" record(s) are monitored (at a user configurable interval) by the hot-standby system. The hot-standby system does not monitor the system or netperfmonservice on the primary or secondary poller systems. If the the poller "control" record fails to be updated, the hot-standby system will take over and begin polling the devices controlled by the failed polling engine. When the failed poller is restored and begins to update its "control" record, the hot-standby system will return to a standby/idle state.

We have recently purchased the hot-standby pack but haven't had the opportunity to install it yet. We are currently working on building the server. We have one problem poller, for whatever reason, frequently (every day or so) stops updating statistics gathered via SNMP polling and syncing with the database but the service continues to update ICMP statistics. So, we hope this helps resolve our loss of statistics during the period the poller is lost. We're working with SW in troubleshooting this problem.

Hope this helps.

chris.conway

rgward has explained this well.

The Hot-Standby will be in an idle state monitoring the last update time of each polling engine in the database - whether it is the main Orion server or an addtional engine.

As soon as the update time doesn't change the Hot-Standby will take over the monitoring for whichever engine fails. When the failed engine starts to update again, the Hot-Standby will go back to its idle state.

A SQL cluster will get round the problem of the SQL server failing.

For the website, this is included with the Hot-Standby, you just need to make sure any maps, custom reports or any of your own resource files are copied across.

You could also have an additional web server license if you wanted to on a completely seperate server that does no polling at all. Again, files would need to be copied across.

Hope that helps

Chris

Analyser Sales Ltd
SolarWinds Premier Partner
www.asl-solarwinds.co.uk

jonchill

Thanks everyone for your feedback and help.

Jon

IT Infrastructure Manager
Pilgrim Hospital
UK

inwsld

Hi Chris,

Could you tell me if the "update time" is configurable? I'm having a problem where my primary polling engine is failing over for reason(s) unknown. I suspect root cause is that my "update time" is too short for the size of the database (about 3600 objects). I'm theorizing that my pollers are sending the updates to my SQL box in a timely manner, but the SQL box may be queueing them just long enough for the update interval time to expire thus causing my hot-standby to take over. Interestingly, the problem goes away after I restart the Solarwinds (SW) services on all three SW servers (primary, secondary and standby) and the SQL service on the SQL box. However, the problem always comes back five to seven days afterwards, thus making me think that I may also have some kind of memory leak on the SQL box (although I've yet to see any problems that way after checking the SQL box).

I thank you very much for your time and assistance.