Dear Davide,

Thanks for your input! To check that I'm certainly not doing anything  
else time-consuming and that it's not a faulty hard drive with a bad  
cluster in mailusers.tab, I made this little shell script:

===
date
grep \"test\.domain\.tld\" ../mailusers.tab
date
../CtrlClnt -s localhost -u ******* -p ********** userlist  
test.domain.tld
date
grep \"test\.domain\.tld\" ../mailusers.tab
date
===

This is the result:
===
Sun Jun 15 15:06:15 CEST 2008
"test.domain.tld"       "1"     "54"    21574   "1"     "U"
"test.domain.tld"       "2"     "57"    21575   "2"     "U"
"test.domain.tld"       "3"     "56"    21576   "3"     "U"
"test.domain.tld"       "4"     "51"    21577   "4"     "U"
"test.domain.tld"       "5"     "50"    21578   "5"     "U"
Sun Jun 15 15:06:15 CEST 2008
"test.domain.tld"       "1"     "1"     "U"
"test.domain.tld"       "2"     "2"     "U"
"test.domain.tld"       "3"     "3"     "U"
"test.domain.tld"       "4"     "4"     "U"
"test.domain.tld"       "5"     "5"     "U"
Sun Jun 15 15:06:30 CEST 2008
"test.domain.tld"       "1"     "54"    21574   "1"     "U"
"test.domain.tld"       "2"     "57"    21575   "2"     "U"
"test.domain.tld"       "3"     "56"    21576   "3"     "U"
"test.domain.tld"       "4"     "51"    21577   "4"     "U"
"test.domain.tld"       "5"     "50"    21578   "5"     "U"
Sun Jun 15 15:06:30 CEST 2008
===

Getting the userlist via CtrlClnt took about 15 seconds while grep  
needed less then a second.

When getting a full list of all users (not only for one domain), the  
processing seems to stall now and again. It however stalls in  
different locations every next time I run it.

Any ideas of what might be going wrong? Could this be caused by some  
faulty setting somewhere?

Sincerely,
Bart Mortelmans



Op 15-jun-08, om 01:13 heeft Davide Libenzi het volgende geschreven:

> On Fri, 13 Jun 2008, Bart Mortelmans wrote:
>
>> Hi all,
>>
>> When I run the CTRL-command "userlist" requesting users for one  
>> domain,
>> it seems to take an unreasonable amount of time before it returns  
>> (20 to
>> 30 seconds). In between, when I monitor the server with the Linux
>> command "top", I notice that "wa" goes up to over 90%. Normal load
>> average on this machine is around 0, but when running "userlist" this
>> will go up to over 2.
>>
>> This is the same even if the userlist which is returned, should  
>> contain
>> no users or only a small amount.
>>
>> The total size of mailusers.tab is reasonably large (over 1Mb), but I
>> would expect CTRL to at least be able to find the users in the list  
>> as
>> quicly as a simple grep for example?
>> For the record: I'm not requesting the full userlist, I use the  
>> command
>> like this:
>> "userlist"[TAB]"domain"<CR><LF>
>>
>> The complete configuration of XMailserver is a copy/paste from an  
>> other
>> server. I'm seeing this both with XMailserver 1.24 and 1.25.
>>
>> The hard drive of this machine seems to be okay. I also seem to  
>> notice
>> something similar on an other machine, although not as extreme.
>>
>> I looked into many different possibilities to find the cause of this
>> (including deleting tabindex), but am unable to find it. Or is this
>> normal behaviour?
>
> Userlist should be very fast. The time of copying out the mailusers  
> file
> (that for 1MB size should be like 0.1s) and the time of reading each  
> line.
> Are you sure you're not issuing other commands, like, for example, the
> ones to query the mailbox/spool?
> Disk activity should be almost NIL for userlist.
>
>
>
> - Davide
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe xmail" in
> the body of a message to [EMAIL PROTECTED]
> For general help: send the line "help" in the body of a message to
> [EMAIL PROTECTED]
>

-
To unsubscribe from this list: send the line "unsubscribe xmail" in
the body of a message to [EMAIL PROTECTED]
For general help: send the line "help" in the body of a message to
[EMAIL PROTECTED]

Reply via email to