Multible Problems receiving and sending E-Mails

[JbS]

sorry, wasn't trying not to answer:
What hard drive setup? raid controller?, write cache enabled or not, disks? -->
hard drive setup: 2 disks 72GB mirrored
raid controller --> don't know --> standard ATA RAID Controler from HP that comes with ML110 (age 2years)
write cache enabled --> don't know

disks --> ATA 72GB 24*7h, server disks
To find out waht RAID-Controller we have I would have to reboot the system, but I'm running our www-page in an other virtualistaions, so I can't just turn of the system.

Does that information help you?
What next??

 

[tom]

sorry, wasn't trying not to answer:
What hard drive setup? raid controller?, write cache enabled or not, disks? -->
hard drive setup: 2 disks 72GB mirrored
raid controller --> don't know --> standard ATA RAID Controler from HP that comes with ML110 (age 2years)
write cache enabled --> don't know

disks --> ATA 72GB 24*7h, server disks
To find out waht RAID-Controller we have I would have to reboot the system, but I'm running our www-page in an other virtualistaions, so I can't just turn of the system.

Does that information help you?
What next??

Hi,

There are a lot of different ML110 servers around. based on the proxperf results I cannot imagine that there is a "real" hardware raid controller in your ML110 - or the other guests take too much disk performance - anyway, you need a reasonable fast environment to run Proxmox and it looks your current machine is not suitable for the load.

If you run virtualized and especially full virtualized like VMware, you should use "premium" hardware in kind of storage:

• Real Hardware Raid Controller with write cache
• SAS/SCSI harddrives with 15 or 10 rpm

eg. a Proxmox Certified Server like HP DL380G5.

 

[JbS]

Hi Tom
Thank you, I thought we had discover that the moment I had sent you the proxperf values.
we will build up new server as we expect more load to come.
And at a down time I will check what RAID-controller we have.
The VR for the www-server for sure is not making allot of IO. We have like 30 Hits a day!!!

So let's get back to the work around:
PC has been started again in the dmz with he's own IP. So new e-mails are still going to the old (bad hw) server. Now there 73 mails in the Queue from the test yesterday. Theoreticly the PC should be relaying the mails now. Can I force it to do so... I would like to get a log or error or something, so we can find out why it's not working!

 

[tom]

Hi Tom
Thank you, I thought we had discover that the moment I had sent you the proxperf values.
we will build up new server as we expect more load to come.
And at a down time I will check what RAID-controller we have.
The VR for the www-server for sure is not making allot of IO. We have like 30 Hits a day!!!

So let's get back to the work around:
PC has been started again in the dmz with he's own IP. So new e-mails are still going to the old (bad hw) server. Now there 73 mails in the Queue from the test yesterday. Theoreticly the PC should be relaying the mails now. Can I force it to do so... I would like to get a log or error or something, so we can find out why it's not working!

hi,
you can flush the mail queue (just click on the arrow and select "flush queue". you can also click on the queue detail so you can see the log why the emails are in the queue.

 

[JbS]

Ok --> flush

First of all --> some e-mails going out worked after starting up the pc, but not all
2nd --> relaying still doesn't work. Getting error:
SMTP RESULT: delivery temporarily suspended: conversation with mail.projectcare.ch[213.158.158.214] timed out while receiving the initial server greeting

after flush I also got Can't resolf host blabla (A) --> checked on the server console and DNS, ping and snlookup works fine.

I booted and now I get the above error again (time out)

Funy. since I have taken over the configuration from the old server I can't connect by ssh anymore??

Maybe I should simply try setting up a server on scratch. Woulden't take long. Maybe it's just a conf error?

Any tipps?

 

[tom]

Ok --> flush

First of all --> some e-mails going out worked after starting up the pc, but not all
2nd --> relaying still doesn't work. Getting error:
SMTP RESULT: delivery temporarily suspended: conversation with mail.projectcare.ch[213.158.158.214] timed out while receiving the initial server greeting

after flush I also got Can't resolf host blabla (A) --> checked on the server console and DNS, ping and snlookup works fine.

I booted and now I get the above error again (time out)

Funy. since I have taken over the configuration from the old server I can't connect by ssh anymore??

Maybe I should simply try setting up a server on scratch. Woulden't take long. Maybe it's just a conf error?

Any tipps?

it looks that you have a network problem or something went wrong during your restore. if possible, reinstall, run proxperf to make sure that your hardware works as expected and reconfigure (without the use of a backup)

 

[dietmar]

You can try to use telnet to debug the problem.

> telnet 213.158.158.214 25

- Dietmar

 

[JbS]

Hi Tom
Ok, I'll set it up again and manualy add the configuration. If that works I'll switch the travic to the PC and will set up the VM the same way. At that point I will be able to once plan some down time and check the raid-controller settings the disks and the filesystem.
Maybe will find something.

But first I need to dave the e-mails on the pc which are waiting in the queue. This are productive e-mails!
How can I same them? are they waiting in a directory??

 

[JbS]

Tom, there is one more thing we need to look at:
- Why do I receive every e-mail from your support system, telling me you updated the Thread, but I don't get other e-mail from proxmox.
(infact I just found a an e-mail from Martin Maurer closing y case, asking for a new licence key. Didn't get the key, as the mail never arrived, but i cot the system-e-mail that he closed the call)

this is the hole mystery - 99% of the e-mails arrive
1% depending on what/which/whos e-mail-Server sends it does not even finish a handshake with us.

Please CHECK the smtp log sending the 'Thread update-email' and compare it to your other proxmox-e-mail-server.
if we know the diffrenz we maybe can figure out what's going wrong!

PS: I'll set up the system anyway

 

[JbS]

Hi Dietmar
If you are working on this ticket today, can you please check on the phenomena I described in my last post:
- Mail from your support-tool arrive
- mail from persons from proxmox to us your e-mail server gets a time out

It's important we track this effect, as it might leed us to the problem on our system

 

[dietmar]

When the logs show you cant connect simply test that, and find out whats wrong:

> telnet 213.158.158.214 25

You tested that? What results? Please solve your problems step by step.

- Dietmar

 

[JbS]

Hi Dietmar
Telnet works fine.
The Server answers with Hello I'm server blablabla

I can enter commands like hallo ello, server doesn't recognice it, but it's happy.

 

[dietmar]

Hi Dietmar
Telnet works fine.
The Server answers with Hello I'm server blablabla

I can enter commands like hallo ello, server doesn't recognice it, but it's happy.

And flushing the queue on the same server still produces the error (timeout) for that host?

- Dietmar

 

[JbS]

yes, but please check on your logs in Proxmox now
Why does one work and the other not?

 

[dietmar]

Hi Dietmar
If you are working on this ticket today, can you please check on the phenomena I described in my last post:
- Mail from your support-tool arrive
- mail from persons from proxmox to us your e-mail server gets a time out

That is simply wrong. Our logs indicates that all mails gets timouts, even those from the support tool (its random - sometimes it work, but most times we get a timeout at your server)

We send all mail from the same server.

- Dietmar

 

[JbS]

Hi there
that's strange. Up to now I have received every e-mail from your system after you (or Tom) have updated the thread!

That's why I concludet that there might be a diffrence in communication.

To bad, that would have been a good lead :-(

Update:
As there might be a Network problem there are only two components between Internet and Proxmox-Gateway: Firewall and a small Switch.
This weekend I exchanged the Switch --> don't think that made a diffrence, but feal free to send me an e-mail!
Firewall was rebooted, but can't imagen it beeing a problem. It's a Zywall 5.

Second test with PC:
After exchanging the Switch I also switched the Port-Reroute to the PC. I had reinstalled and manualy configured Proxmox beforhand.
I get the same problem as I did while the first test: It can't relay to the other domains. Settings all look good, but it times out. Telnet is fine.

As SSH works on that machine I would like to have you connect to it and help me find the problem. OK?
Can you call me and I open up SSH for you?

 

[JbS]

Is there a way of looking at the hole server-log (messages or syslog) from the web-console?
since the new version we have the message tracker, which is great!!!
But I'm missing a general sys-log viewer as we had beforhand.

 

[dietmar]

Hi there
that's strange. Up to now I have received every e-mail from your system after you (or Tom) have updated the thread!

That's why I concludet that there might be a diffrence in communication.

To bad, that would have been a good lead :-(

OK, you receive mails from our support forum - i was able to reproduce that. (sorry for the confusion, but the forum and support system are different, so we talked about different things).

Update:
As there might be a Network problem there are only two components between Internet and Proxmox-Gateway: Firewall and a small Switch.
This weekend I exchanged the Switch --> don't think that made a diffrence, but feal free to send me an e-mail!
Firewall was rebooted, but can't imagen it beeing a problem. It's a Zywall 5.

There are various posting about problems with Zywall 5. Whats your firmeware version? Please update to the newest version an test again.

- Dietmar

 

[JbS]

Hi there
The Firmware is V4.00(XD.7). Not the newest. I can Upgrade it tonight, but not right now, as it's a productive system.

Let's remember that the e-mail problem started about 7 weeks ago after upgrading to Gateway 2.1 the same firewall has been installed for 2y now. now Update no, changes as we believe in 'don't touch a running system'.

But I'll be happy to update the firewall - wich meens taking a risk that it won't work after the upgrade...and down grade is not a good idea!

In the meentime:
Can we make a phone call this afternoon and look at the PC by ssh?
I think it might help finding the problem, what ever it is.
Please..

 

[JbS]

Good to hear that we do have two diffrent systems:
what is the diffrenz between the normal e-mail-system and the thread updates?
Do the logs give any hint?

My experiance is that certain e-mail servers have a problem sending uns e-mails, others just work fine.
So I sort of dout a network problem, more seems like a protokoll-problem.
That's why I would like you to look at the pc i set up. Even better would be the productiv system, but ssh doesn't work anymore.