Files Changes Not Detected

terrihawkins · Guest

I am using Vice Versa 2.5 and VV Enginer 1.1(Build 7). I have several profiles created and in the scheduler tell it to "Monitor Source and run when file changes are detected". Most of the time, changes are detected and the profile runs, but frequently it does not detect them. If I run the job manually, it finds them and replicates it over to the other server. Any idea why this would happen? I can't find any consistency to it, it will keep working then suddenly stop. One profile may continue to work while another one stops. I have found that stopping and restarting the VV engine will help and everything will run smoothly for a few days, but then it starts again. Any insight is appreciated.

TGRMN Software · Site Admin Joined: 10 Jan 2005 Posts: 8763

Hello

are you monitoring network folders or local folders? Where does this problem happen, network folders or local folders too?

thanks
_________________
--
TGRMN Software Support
http://www.tgrmn.com
http://www.compareandmerge.com

terri8502 · Joined: 27 Aug 2008 Posts: 21

I have ViceVersa on a central server and am monitoring folders and copying changes to two other servers.

TGRMN Software · Site Admin Joined: 10 Jan 2005 Posts: 8763

Are the monitored folders on the central server (local) or on the other two servers (remote)? thanks
_________________
--
TGRMN Software Support
http://www.tgrmn.com
http://www.compareandmerge.com

terri8502 · Joined: 27 Aug 2008 Posts: 21

The source and target are different from the ViceVersa server. Sorry, guess I didn't make that too clear.

TGRMN Software · Site Admin Joined: 10 Jan 2005 Posts: 8763

I wonder if a momentary gap in network connectivity could be the issue. If you disable and re-enable the profile does monitoring restart? Are you using the latest VVEngine 1.1 build 7 ? thanks
_________________
--
TGRMN Software Support
http://www.tgrmn.com
http://www.compareandmerge.com

terri8502 · Joined: 27 Aug 2008 Posts: 21

According to our network guys, we NEVER have any connectivity issues. Smile

All kidding aside, yes, we are using the most current version of ViceVersa and the VV engine (I did an update test on both products and updated Vice Versa). One of the reasons I upgraded was to see if it would eliminate this problem because I thought I read there was an improvement in monitoring deep folders. Many of ours are multiple layers deep, but we also have issues with some that are only 4 levels down.

There very well could be network issues, but I have in the past seen in the log where the connection is lost but the profile does not need to be reenabled and it just continues on it's way.

thanks
terri

terri8502 · Joined: 27 Aug 2008 Posts: 21

Ok, not sure if anyone cares or not, but here is what I have found concerning the file changes not being detected. This is just an observation made from consistent error messages.

We are running ViceVersa (the newest ones) against webservers. We are replicating the content from one web server to another. It appears that if ViceVersa is running at the same time as an application pool recycles (source or target), the job breaks and the file detection stops. This seems to be the case because we starting kicking out emails for each job if it errored out. The time on the email fits exactly to the time on our web server event logs when the app pools are recycling.

I have temporarily eliminated the file detection problem by also scheduling the vice versa job to run at scheduled times (which is not very efficient for us, but at least it doesn't stop running, which is worse). Unfortunately, right now I don't see a way to tell it to stop the schedule at a certain time (say OOD) and pick back up in the morning, which would stop the job from running during periods when no one is doing any web updates. Is there a way to do that? Also, do you think my hypothesis about the app pool is correct?

TGRMN Software · Site Admin Joined: 10 Jan 2005 Posts: 8763

Hello,

what happens during application pool recycles ?

thanks
_________________
--
TGRMN Software Support
http://www.tgrmn.com
http://www.compareandmerge.com

terri8502 · Joined: 27 Aug 2008 Posts: 21

Application pools provide a way to isolate Web sites and web applications from each other even though they are being hosted on a common server. Each application pool is given its own set of server resources and can be configured with different identities. If a Web site crashes, it won’t effect sites in other application pools.

When you recycle an application pool, HTTP.SYS holds onto the client connection in kernel mode while the user mode worker process recycles. After the process recycle, HTTP.SYS transparently routes the new requests to the new worker process. Thus, the client never loses all connectivity to the server - the TCP connection is never lost - and never notices the process recycle.

Because of memory leaks problems (which we are working to correct), we have an aggressive schedule for application pool recycling, including when the memory gets to a certain level and at scheduled times when we know bots and our own reports are hitting the server (early in the AM).

The emails that VV kicks out say the source or target has become unaccessible, and the job fails. The time on the emails coinside with application pool recycles (the scheduled ones, I have not had VV running during a memory based recycle yet). These are mostly our late night recycles. I plan to alter the time on the VV jobs by a few minutes so they will kick off after a planned recycle is complete to see if that will make a difference.

I am not sure why this would have an effect on ViceVersa, or if it does at all. It could just be a coincidence.