Quick Help anyone? All FTPs have stopped

[clk430]

This happened last week. No FTP maps were able to get or send. We rebooted and bounced the service, several times, but this did nothing.

It is happening now and we are 2000 files behind. Network and firewall folks confirm no activity, however they don't see a problem. Any ideas?

 

[clk430]

OK, 2 hours later, all FTP's just started by themsleves. No errors, traces etc. since all maps just "hung." Thank God for overtime.

 

[janhes]

Maybe it was a problem at the other end and you don't have a timeout on your connection.

 

[BocaBurger]

When you see a "hang" in UNIX, do three things: Get a snapshot from the Event Server Monitor Tool, Get a Snapshot from the Management Console, run the ui script against the process ID three times about three minutes apart. Send all to support.

On Windows, do 1 & 2 and pray.

Sounds like you had hit a resource limit, or there was a conflict that locked thinsg up (DB update) along with reaching maxthreads or something like this.



BocaBurger
<===========================||////////////////|0
The pen is mightier than the sword, but the sword hurts more!

 

[janhes]

Did you try a manual ftp connection to check all was well with the link?

 

[clk430]

I tired the snap shot in the event server monitor tool, and the files errored everytime they opened. The snap shot on the management console would have just showed all FTP inbound and out bound maps just sitting there.

The worst part is that all the network and firewall people said nothing on their end was wrong. I tired to manually connect with our FTP tool, and all connections were made.

I did drop maxthreads down to 10 a few weeks ago - thus eliminating GPF's, but perhaps that could have caused a resource limit?

 

[clk430]

Also, Boca, you mentioned a BD lock-up. We do occasionally get a DB error in launcher.txt.

*** ERROR:

System: X12 Inbound
Component: db_writer
Map: E:\Mercator6.7\cmgr\mmgr\db_writer.mmc
Start Time: Wed Mar 08 08:31:29 2006
Instance: 635
Run Time: 0.0007 seconds
Return Code: 12 - Source not available

Input 1: File: E:\Mercator6.7\cmgr\mmgr\audit\x12\cop\x12_sortx12.cop
Output 1: Database: 0008
Output 2: Database: 016B
Output 3: Database: 02A1
Output 4: Database: 0450
Output 5: Database: 06F0
Output 6: Database: 08DE
Output 7: Database: 0A62
Output 8: Sink: update_rc

Audit File: <none>

==============================================================================

But things seem to process fine.

 

[BocaBurger]

That is not a DB error. It is a file source error.

When you reduced maxthreads, did you also set WatchMaxthreads below that value?

Check the file size of the .mss file, if zero it will error, the tool needs to be connected to a running Event Server and some maps have to run, to get a snapshot. This tool is basically unchanged since 5.0 days and works OK. The MC snapshot contains the history and connections information too.
Snapshot, not screen shot.



BocaBurger
<===========================||////////////////|0
The pen is mightier than the sword, but the sword hurts more!