Sendmail Performance tuning

I have default configuration of sendmail(8.13.8) on centos 5.

I am able to send 1 lack mails in 1 hour. i saw some articles where they are saying, they are able to send 3 to 5 lacks emails in 1 hour.

How can I tune the sendmail to send those many emails.

sasidhar p.
Who is Participating?

Improve company productivity with a Business Account.Sign Up

robocatConnect With a Mentor Commented:
The new server should perform better indeed, so something is very wrong.

You really need somebody who knows about Dell raid systems to take a look at this because this can be tricky to troubleshoot. There's a lot of things that can be wrong: using non-Dell disks with a Dell raid controller, using the wrong driver in centos ... Perhaps there's a networking problem?

If possible, use the previous sendmail configuration to rule out that this is the cause of the problem.

Just a remark: currently you have raid zero which offers no protection against disk failure, which is no worse than your previous setup, but you need to determine if that is acceptable for your situation.
What is a "lack" e-mail?

In general, high volume sendmail servers are limited by the perfomance of the disk IO of the server. You need a lot of IOPS, so a good RAID system with lots of disks and lots of cache will increase the throughput. Or a high performance SAN.
sasidhar1229Author Commented:
What is a "lack" e-mail?

Sorry I misspelled it. That is Lakh.
Build your data science skills into a career

Are you ready to take your data science career to the next step, or break into data science? With Springboard’s Data Science Career Track, you’ll master data science topics, have personalized career guidance, weekly calls with a data science expert, and a job guarantee.

Same question: what exactly do you mean with "lakh e-mail" ?
Jan SpringerCommented:
A standard sendmail configuration should be able to do what you need as long as the number of child processes spawned are sufficient (you would have to intentionally restrict that), you have fast disk i/o and uncongested bandwidth leaving the server.
sasidhar1229Author Commented:
One hundred thousand (100,000). We call it lakh (1,00,000).

Presently I am able to send 100,000 e-mails per hour.
What's your hardware (server) setup? RAM, CPU, disk, ...
I guess thats is spamming (I apologize - if I am wrong)

As for performance tuning - you might want to check for multiple threads.

As I understand - If send rate is that high (and I have seen that - First hand) - Majority Email providers ban this behaviour unless you are certified big organization like Google/Microsoft/IBM/Intel etc - Stress is on reputable organizations.

I am pretty sure if you check your mail log - you will have loads of delayed /mail bounces - Which ultimately will result in IP Ban/Domain Ban -  Not to miss out on Spam black lists.
sasidhar1229Author Commented:
Motherboard  :   Dell Single socket 1067FSB - Quad Core Capable PowerEdge 840
Processor       :    Intel 2.4 GHz 1066FSB - Conroe Xeon 3060 (Dual Core)
RAM              :     OEM Standard DDR2 667 ECC [4096MB]
Drive Controller : Generic \ Onboard \ SATA
Hard Drive 1       :   Western Digital 250GB:SATA2:7200RPM WD2500JS [250GB]
Your hard disk is almost certainly your bottleneck. A single SATA disk can process 100 IOPS at best.  SATA is not suitable for such a task.

Moving to 15000RPM SAS disks will double your performance.

Moving to a good (server type) raid controller with plently of cache (battery backup write cache) and at least 5 SAS disks in RAID will increase your performance to what you desire.
sasidhar1229Author Commented:
Hi robocat,

We upgraded our server. In a strange way the performance decreased by half. Earlier we used to send 10000 mails in 6 or 7 minutes. After upgrade for the same 10000 mails it's taking 15 to 16 minutes.

This is the hardware configuration.

      Os changed from centos 5.0 32 bit to centos 5.9 64bit.

        Motherboard               Dell Single socket 1067FSB - Quad Core Capable PowerEdge 840

        Processor               Intel 2.4 GHz 1066FSB - Conroe Xeon 3060 (Dual Core)

        RAM slot 1               OEM Standard DDR2 667 ECC [1024MB]
        RAM slot 2               OEM Standard DDR2 667 ECC [1024MB]
        RAM slot 3               OEM Standard DDR2 667 ECC [2048MB]
        RAM slot 4               OEM Standard DDR2 667 ECC [2048MB]

        Drive Controller               Dell \ 9G Drive Controller - SAS/SATA - budget raid \ SAS6iR

        Hard Drive 1               Generic 146GB:SAS:15kRPM 146GB15KSAS [146GB]
        Hard Drive 2               Generic 146GB:SAS:15kRPM 146GB15KSAS [146GB]
So everything stayed the same except the raid controller and centos version?

You will have to troubleshoot your configuration for IO performance problems. Too bad you changed two things at once (raid and OS) because this makes it harder to determine the source of your problem.

Check the RAID configuration is ok, and that the most recent 6iR driver is installed in centos. The SAS drives are hopefully Dell certified ?

Also check that the sendmail config hasn't changed in any way.
sasidhar1229Author Commented:
We tried to change the complete hardware. Present machines IP reputation is very good. We asked for the new machine with the same IP, but the hosting providers said they can't keep the same IP for the new machine. With the new IP we can't send e-mails what we expected to.

Two warm up the IP itself takes some time.

Can we reload the os to earlier version?
And what is the good RAID configuration for our requirement?

I am confused that, RPM and Raid both are better than older configuration. It must work at least or better than the older one.
sasidhar1229Author Commented:
This is the raid information

[root@ ~]# omreport storage vdisk controller=0

Virtual Disk 0 on Controller SAS 6/iR Integrated (Slot 1)

Controller SAS 6/iR Integrated (Slot 1)
ID : 0
Status : Ok
Name : Virtual Disk 0
State : Ready
Hot Spare Policy violated : Not Applicable
Encrypted : Not Applicable
Layout : RAID-0
Size : 272.25 GB (292326211584 bytes)
Associated Fluid Cache State : Not Applicable
Device Name : /dev/sda
Bus Protocol : SAS
Media : HDD
Read Policy : Not Applicable
Write Policy : Not Applicable
Cache Policy : Not Applicable
Stripe Element Size : 64 KB
Disk Cache Policy : Disabled
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.