Looking for a file copying tool with alerting

D

Dale_tem

Suspended / Banned
Messages
2,839
Name
Dale
Edit My Images
Yes
Hi

Looking for a reliable tool to copy a large number of files from one server to another (has to use volume shadow copy). I would like to run periodically (30 min intervals) and alert me of any errors via email.

Thanks in advance.
 
You could just setup a scheduled task running an xcopy script with a few options setup.
I think you could then use an smtp mailer issue to sent the result via your smtp service.
 
I use either Robocopy CLI or Richcopy for our server migrations. Call me old-school but they are tried and tested, fast and robust with extensive capabilities.

And I hope Andy's not referring to OSX as a proper operating system in terms of server? ;)
 
When you say 'server' are we talking Windows Server or desktop OS? If Windows Server then I'd have a look at DFS (Distributed File System). If Windows desktop OS I'd second robocopy - brilliant for server migrations, etc. and will output to a log file. BackupAssist does what it says on the tin and from memory will provide an email report but obviously is a cost option
 
VSS is used for backups.
So what you are talking about is backing up files from one server, to another?

My First thought was DFS-R. If you are dealing with/need servers, then chances are you should have some decent monitoring in place. You can use a Nagios check that warns you if your backup goes stale. Otherwise, a PowerShell one liner run as a scheduled task orta do it.

if (!(gci c:\ | Where {$_.LastWriteTime -gt (Get-Date).AddMinutes(-30) })){ Send-MailMessage -From "afasoas <afasoas@blah.blah>" -To "afasoas <afasoas@blah.blah>" -Subject "Backup Stale" -Body "Backup is dodgy?" -SmtpServer "mail.blah.blah" }
 
Neilc28 said:
I use either Robocopy CLI or Richcopy for our server migrations. Call me old-school but they are tried and tested, fast and robust with extensive capabilities.

And I hope Andy's not referring to OSX as a proper operating system in terms of server? ;)
Click to expand...
Mines been up for since the last power outage. Nothing wrong with it as a server operating system.

Then again neither with a well configured <fill with any operating system name>.
 
One of my Linux servers at work has a current uptime of over 10 years. An incremental backup with an email when it fails is relatively simple.
 
Thanks for all the info. The solution is for Windows servers in a data centre and will be a short term solution till we move to SANs using object based storage via the S3 protocol. At that point backup is all taken care of.

No chance of using anything other than Windows as we are using IIS/ASP/SQL2014 setup. The files are accessed by the website and we are talking of around 600gb within the first year which need replicating to the backup server on cold standby.

The recommendation of DFS I will investigate. We monitor via a number of methods already, I just want a fire and forget system that will email me in the event of an issue and I can check is correct in a weekly basis.
 
Not to query the deployment but would a failover cluster configuration not be more effective?

The main concern I'd have with the replication is if the SQL logs and database don't match up perfectly, could be a pain to resolve when you urgently need the standby server operational.

I now understand why you wanted VSS capable replication software!
 
The reason we are not going to cluster is the migration to SAN based storage later on and how long that would take. We will start with 2 DB servers, one replicating to another. When we get the SANs up and running, the main DB servers will use the SANs and the back up DB server will become a main DB server also accessing the SANs (software designed to split customers across DB servers). We will have a 3rd server as a standby for the DB servers (expecting 5 DB servers or more over the next 3 years). SQL will use log shipping on 15 minute intervals and the SANs will replicate between data centres automatically.

We are having our authentication servers in an SQL failover cluster.
 
We use DFS-R for SQL Transaction LogShipping between two DCs

We have direct access storage for each SQL server in one DC (mirroring setup for HA)
We have a SAN for a SQL server cluster on our DR site

The DAS gives us nice low latency < 5 ms
The SAN does give us incredible IOPs but but much higher latency > 100ms

Just wondering what you're storing that requires sharding your data? If it's for performance, then the SAN might set you back a bit.

We're actually trying to ditch SQL Server, as sharding to give dual master is a PITA and use Cassandra for write optimised work and then PostgreSQL for everything else - mainly for data our services cache.
 
afasoas said:
We use DFS-R for SQL Transaction LogShipping between two DCs

We have direct access storage for each SQL server in one DC (mirroring setup for HA)
We have a SAN for a SQL server cluster on our DR site

The DAS gives us nice low latency < 5 ms
The SAN does give us incredible IOPs but but much higher latency > 100ms

Just wondering what you're storing that requires sharding your data? If it's for performance/managing load, then the SAN might set you back a bit.

We're actually trying to ditch SQL Server, as sharding to give dual master is a PITA and use Cassandra for write optimised work and then PostgreSQL for everything else - mainly for data our services cache.
Click to expand...
 
Bum, dual post in error. 'Edit' ! 'Reply'

On further edit, using two mirrored SQL Servers with their own DAS in our live DC is obviously good for HA.

Clustering, AFAIK (I'm not a DBA, just mild involvement) relies on shared storage i.e SAN between the clustered nodes, so in our case the SAN is SPOF. It's a fair compromise in our DR environment as we planning to step away from SQL Server soon(ish) but I'd be reticent to use it in production.
 
Last edited:
Sorry for repeatedly posting.
If you are using Nagios for your monitoring, I have a nagios/powershell script that montors DFS backlog counts - it's a useful early warning system. If you are not using Nagios, it could be tweaked to run as a scheduled task and generate email alerts.
We also have replication checks too I could dig up for you, if needed.
 
Last edited:
You must log in or register to reply here.
Back
Top