MXe 3310 issue

[BenM]

Hi,

Our MXe is giving trouble, was wondering if anyone can help.....

System Description:
MXe 3310 on its own WAN, Clustered with 2 other MXe 3310s. WAN Link 100Mbit to hub MXe synchronous. MCD Software rev 4.2
Handsets all 5320/5330 PoE

Symptoms:
Box is pingable

Incoming, Outgoing and Extension - Extension calls work fine

Web management interface is very slow (30-60 minutes to login) or non responsive, usually the latter.

Yesterday (when maintainer was here) on the RTC repeated  ps showed the CallEngine process running at >80% cpu which, for a *nix process, is bad news and explains the lack of responsiveness from the system.

When a user logs out of a handset (all users are hot desk) the handset will let them log in, however keys are not displayed and we get a "please wait while upgrades are being performed..." message on the handset; the handset does get the right DN and functions as expected.

Can't connect to MXe via Telnet  (port 2002) (login fail for valid logins) or the web interface (times out) so no visibility of logs. My serial lead seems to disagree with the serial port on the MXe.

The other two MXes in the cluster are complaining (understandably) about SDS replication errors

Things we have tried:
issued a reboot command from the RTC - everything seemed OK until maintainer started a backup; backup hung on "Multiline device key assignments" according to the logging on the RTC.

So the maintainer says replace the HDD or the System... obviously I am reluctant to do either if there is another fix possible... for example are there any RTC commands that can check out the HDD state (like fsck) on a running system just in case the Multiline Handset Key definition database is corrupted by bad sectors or some other issue?

TIA

BenM

[ralph]

The only command I have that may show anything is 'ataShow' from the RTC shell.

I have see similar things that a reboot resolved.
Of course I'm concerned about the back up but at this point you may be forced to reboot anyhow to see what turns up.
I suspect you'll be fine.

Did your program reboot actually go through and reboot the system?
If not, what you may want to do is have your maintainer be on site with a spare hard drive (and latest backup you do have) and reboot the system with the reset button. - not command line or power cycle.   (Fingers crossed) it will come up just fine and then you can get your backup.

Ralph

[BenM]

yep - the program reboot went just fine. no errors (other than normal ones about network voicemail not being licensed - well doh! I guess that is because it isn't!)

The last change on the system was on Wednesday this week, the one before that was in May, after which we took a backup - the machine has worked fine since then until yesterday so the current backup we have is complete except for the change made this week (adding another subscriber)

Regards

B.

[ralph]

You may have to have your maintainer go directly to Mitel tech support to help find what is chewing up your processing power.
Hopefully your software is 9.0 or above.

At one time I have seen where a SNMP monitoring tool created such a problem.   Are you using something like WhatsUp Gold?

Ralph

[BenM]

We do look at what the switch thinks its network card is doing using PRTG but I would have thought that any issues caused by that would have manifested way before now (We installed the systems in January 2011) and anyway be cleared by a reboot...

The maintainers are a small company here in the UK...  called BT.... you may have heard of them and I suspect that their standard fix process will cut in so that we will end up with a new box which doesn't actually fix the problem; I assume this will save them corporate 'face' though not with us of course.

Yep - our Software is MCD 4.2 so well in advance of 9 (I never could understand the mad version numbers of the Mitel software) its a bit of a blessing really because we can disconnect the MXe from the network and have all the phones fail over to another site. Tried that yesterday and it worked just fine

B.