On occasion, users will receive 'An unknown error (mail.TRY_AGAIN) has occurred. on their web client... Below is the log info:+

2011-06-27 18:34:25,823 INFO [btpool0-42239://email.domain.com/service/soap/GetMsgRequest] [name=vanessa.lucero@domain.com;mid=731;ip=10.1.10. 60;ua=ZimbraWebClient - IE6 (Win)/6.0.7_GA_2470.RHEL5_64;] soap - GetMsgRequest
2011-06-27 18:34:26,098 INFO [btpool0-42245://email.domain.com/service/soap/SendMsgRequest] [name=wilburn.anderson@domain.com;mid=1275;ip=10.1. 10.60;ua=ZimbraWebClient - IE8 (Win)/6.0.7_GA_2470.RHEL5_64;] SoapEngine - handler exception
com.zimbra.cs.mailbox.MailServiceException: try again: message send already in progress: 1309213939546
at com.zimbra.cs.mailbox.MailServiceException.TRY_AGA IN(MailServiceException.java:471)

The server doesnt seem to be under any exceptional load at this time. Our configuration for 4000 users is as follows:

proxy server
2 mailstores 16gb ram each 15k fiber drives 10 spindles raid5 for stores
2 LDAP servers
1 archive

These are under vsphere 4.x Mailstores have 2 dedicated cores each, and hover around a load of 2.x with about 1200 web clients active per server at a time

I have noticed every once in a while a 'please wait while contacting server' message in the middle of email.. its rare, however. I have followed the tuning guidelines fairly well.