Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations SkipVought on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

SCO Openserver 5

Status
Not open for further replies.

FLAdmin

MIS
Mar 6, 2006
33
0
0
US
First of all I'm a windows admin with just enough knowledge to get me around Unix a bit for small things. However we run our main application wich is FoxPro for Unix based.

The problem we have is Intermittent Hangs to the point where it takes about 1-2 minutes to see any kind of screen update. My only solution has been rebooting the machine unfortunately. I wish I could give more information but if anyone can point me in the right direction I would be happy.

I would like to know also how to do a chkdsk equivalent type command to check for bad sectors on the disk without harming any data.

Also does this thing have any simple path or order to start troubleshooting, maybe a log? and is there any gotchas to watch out for that are worth knowing. thanks

This Machine is: (incredibly)
3.6 Gz P4
1GB RAM
HD, unsure. I believe 2, one for backups.
 
SCO_SV sco 3.2v5.0.6 Pent4 03/07/2006

14:03:30 freemem freeswp availrmem availsmem (-r)
14:03:35 120060 1520472 128631 308242
14:03:40 120064 1520472 128631 308242
14:03:45 120061 1520472 128631 308238
14:03:50 120045 1520472 128627 308182
14:03:55 120043 1520472 128627 308182

Average 120054 1520472 128629 308217

FYI I believe the Box has 1 GB RAM but i will doublecheck next reboot.
 
The system is already patched as custom said to you:
RS506A: Release Supplement for SCO OpenServer Release 5.0.6 (ver rs
RS506A: Software Manager Supplement (ver rs506a)


Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ181-2886
 
Since VHAND has been listed twice, you may benefit by decreasing some kernel parameters which deal with Disk buffers.
What is the output of these commands:
# sysdef|grep BDFLUSH
# sysdef|grep NAUTOUP

These are parameters which tell the O/S how often to check the disk buffer for pending write operations, and how old a "dirty buffer" should be before it is written back to disk. Your Foxpro application vendor may have recommendations for these. In the older days where CPU and DISK were slow, we achieved better overall performance by letting the buffers hang around for awhile before writing them out to disk. With current hardware, it is often better to check these and flush them to disk often, so each pass only has to write a relatively small amount of information.

Please read and review the information in this link:

It sounds like it may reflect your symptoms. Some of the recommendations might be beyond your comfort level, so proceed with caution.
 
I must be getting old and tired. On that link I provided above, the section which might apply is the last regarding "vhand spinning".
 
I have Olympus tuneup and It looks like POPPER is using up A GIG OF RAM and RISING...!! and the user "Jose" as well as no one else can check their pop mail... how to i tame this beast..??? the system is most likely crawling because of it
 
It looks like it killed itself. after reaching about 1.2GB of ram usage. Now his outlook keeps asking for POP user pass even though it is correct, yet it works perfectly fine for any other user. He is sucsessfully logged into a telnet session to sco with the same user pass right now.

do you know if popper maybe only likes to be checked for mail in certain time intervals or anything?
 
Kill his OUTLOOK (or whatever process on the PC is checking the mail)
Go to /usr/spool/mail
Look for all instances of "jose":
Code:
# ls -al | grep jose
-rw-------   1 jose     mail         485 Mar  7 15:24 .jose.pop
-rw-------   1 jose     sys            0 Mar  7 15:24 jose
(In the example above, this is a very small mailbox)
You are probably bouncing his Email between /usr/spool/mail/jose and /usr/spool/mail/.jose.pop

Once that has settled down, it will just sit in /usr/spool/mail/jose. Check the size. If it's big, you probably have no "EASY" choice but to clear it out.
Other options are quite tedious, or require you to configure some other mail reader program.
If you wish to clear it out, just do this (there is no undelete option here, so make sure you are OK with losing all his pending mail):
# >jose
 
here's the deal... He's the boss. so i gotta be a little careful. However most emails are automated over time to his inbox so he gets well.... hundreds some days. I've seen it sometimes pull 26mb - 50mb into his outlook in a single gulp. luckly i moved all the user emails to an exchange server. I have not however been able to correctly redirect some email. eg. my exchange server relay's email for user@flmiami.com but not user@scosysv@flmiami.com my solution for now is to have 2 accounts in their outlook. one for exchange and the other for pop3 from unix for internally generated emails. and the output goes directly to exchange. nifty when it works.

i gotta think hard about what to do on this one... unfortunately more emails pile up every second.
 
I would recommend you try a product on the PC which will display the Email list without actually pulling it in. This will allow you to selectively delete the ones which may be huge (and hopefully unwanted), yet preserve the ones of importance. I use Mailwasher ( I'm sure there are others. You would still have the problem the first time it connects to the server to check mail, as it launches the popper daemon just like Outlook, and his Email will get copied. But, once that goes through, you may be able to mark the ones you want to delete. Outlook will timeout before it ever completes, which just puts you in this on-going cycle.
If you DO run such a utility, change his OUTLOOK settings to NOT automatically check for incoming mail every x-minutes. Make that a manual process so you don't have 2 client programs competing for the same source mailbox.

You can also use the built-in "mail" command from a UNIX shell (while logged in as "jose"). That would probably be a better option, but the interface is awful. Try if first as a user who doesn't have such a large inbox.

Another option is to install PINE on the SCO box. This has a better user interface that the built-in "mail" command.

Just curious: How big is his pending mail file?
 
okay I trashed his mail file with the ">jose" but he still cant check his pop account. it says its being used by another user... ?
 
By the way his file grew to 1.4 GB in about a day or two. its insane. I ftp'd it out to a windows box and zeroe'd it.
 
The POP account problem will eventually go away. Once again, it's probably in the "copy jose to .jose.pop" or "copy .jose.pop to jose" process. This annoyingly occurs every time his PC tries to look for new mail. Once Outlook gives up, the popper daemon has to copy .jose.pop back to jose. Leave his Outlook OFF and make sure there is no "popper" process running for his user account. Then re-check /usr/spool/mail/jose for size. You can also make sure there are no processes running on that file:
# fuser /usr/spool/mail/jose

Make sure he is the owner of that file. Also check for a /usr/spool/mail/jose.lock (might be jose.lck). This is created as the mail transport tries to add to his mailbox.

You may have to develop a script which will monitor his mailbox and alert you when it gets above a certain size. I'm not an Exchange expert, but I don't think it would be really happy with such large Emails either.

You may still be able to reduce the impact all of this has on the Foxbase users by reducing those Kernal parameters I had mentioned earlier (NAUTOUP and BDFLUSH). If you set each to "1", then the system will chug along updating the disk every second, but with much shorter writes.
 
Is it feasable to just ">jose" the /usr/spool/mail/jose file as well as the .jose.pop file? I'm trying to delete all this. The system has been hanging alot today.

Most notibly when I Opened up startx and checked any email regarding the lp. it just locks upon touching those, so i deleted all the emails there. also as soon as I opened the display adapter choose dialog. it immediately froze
 
well deleting that .pop file did the trick for that. He can check his email now.

however I cant explain why it was over a gigabyte of inbox so quickly since SMTP traffic does not come from the outside to that box...
 
With his permission, grab a copy of his mailbox file when it is of reasonable size and look at the mail headers. You should be able to quickly see where the mail originates.

I think you've identified your system performance problem.
You've passed the Unix SysAdmin initiation test.
 
Thanks man. It's like a cryptic dos. luckily dos 2.0 was where i started. Today I pretty much figured out how mail works, monitoring processes, identifying hogs, Qpopper is evil and I must try not to use it if possible. the hosts file missing server entry delayed emails. thanks guys.

I wrote a windows app to quickly view log files from a unix box with a simple drop down menu to choose different logs and an excel like interface with a single button click. I'll post the link when I unhard code the user/pass and let it use an INI instead so you can add locations of your own logs... its really handy.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top