martes, 9 de julio de 2013

Blank email body on Exchange 2010

Hi people!

Today I would like to point out something that drove me crazy for over a week. A customer called us regarding some messages being generated from a server, let's call it SERVER1, and sent through smtp relay configured in Exchange, were being received with blank content (no email body). At first I though to myself... "Oh my god, here comes another end-user Outlook problem", but after I tried it myself, I saw that no matter what client you tried to open the message with, outlook, gmail, or any other web or installed client, all the emails sent from this server were completely blank.

I tested it myself with the following:

 220 internal.relay.domain.com Microsoft ESMTP MAIL Service ready at Tue, 9 Jul 20  
 13 13:05:46 +0200  
 helo mydomain.com  
 250 internal.relay.docout.es Hello [172.28.0.35]  
 mail from: testemail@relay.com  
 rcpt to: destination@domain.com  
 250 2.1.5 Recipient OK  
 data  
 354 Start mail input; end with <CRLF>.<CRLF>  
 to: destination@domain.com  
 from: testemail@relay.com  
 subject: subject from SERVER1  
 This is my content  
 .  
 250 2.6.0 <5cf64727-f4bf-44ca-a084-e4e4d432964b@SERVER1.domain.com> [Internal  
 Id=3365140] Queued mail for delivery  
 quit  
 221 2.0.0 Service closing transmission channel  

Sure enough, just as customer said, this mail was received with blank contents instead of the "this is my content" text.

viernes, 5 de julio de 2013

VMware & W2K8R2 & Exchange & DataRecovery = Cannot quiesce virtual machine

Today I encountered yet again the intermitent and annoying problem of vms not being correctly backed up in one of our DataRecovery appliances for one of our customer. After analyzing, I wasn't surprised to find that one of the VMs, Windows 2008 R2 with Exchange server 2010 installed, could not be taken a snapshot, therefore failing to backup the vm whatsoever.

I wouldn't normally worry too much when this occurrs, as normally we use DataRecovery only as a "just in case" tool, but all our customers have a mandatory system and file remote backup in place. BUT, with this particular customer, some super-duper-intelligent salesman decided to sell them DataRecovery as a backup solution, which, not only is it crazy and unreliable, but also makes it complicated to manage, as we cannot count on all backups running smoothly every time...

Anyway (enough about complaints), back to the point. If you MUST no matter how have a DR backup for a machine and it's giving you problems, the only way to solve the issue without having to should down the server (because as our friend Murphy know, backup only fails on production servers, never on development) is what our vmware friends tell us on KB 1031200 and 1031298, disabling VSS writers usage on VMware tools.

For my particular case, this is how it went:

1. You detect a really nice message saying your backup is f*?¿ up:

 7/5/2013 9:31:05 AM: Normal backup using 3 Wednesday Bck  
 7/5/2013 9:32:20 AM: Failed to create snapshot for CUSTOMER-EX02, error -3960 ( cannot quiesce virtual machine)