Developer forum

It seems like a disk locking issue. The problem is that the file, which contains no data, cannot be written by the indexer within a 2 second time frame. So if you have something that could also look at that folder, it could also lock it. An iisreset does not solve it?

Hi Nicolai,

I have not tried an IIS reset but that should probably solve it.

What I am trying to understand is how to prevent it from repeating.

The website is hosted on a dedicated VM and there is no collaborative work on the files.

That's why I was suspecting a misconfiguration of some sort. Could there be another scheduled task locking that folder?

Thank you,
Adrian

Not built in. No custom open of indexes, repositories or quires?

Hi Nicolai,

Nope. None.

I have noticed though that we had very little space available on the server. Could it be related?

Adrian

Yes - of course!

Hi Nicolai,

I keep having this type of error on the same solution (currently running on 9.9.6) and I have checked the space available and it is not an issue. I have plenty of space on the server.
What other issue can there be?

On the filesystem I don;t see any LOCK file:

If I add one manually, the error changes but it won't Resume indexing.

Adrian

It happens if 2 instances are trying to run at the same time. Maybe a schedule service on another server calling the same URL?

It is not an Azure web app?

Hi Nicolai,

I have just one scheduled task that runs every hour.

The solution is hosted by the customer in a private cloud (it's actually a VM).

Adrian

Hi Adrian

I do not know then - either 2 tasks is running at the same time or something happens on the server that will lock the lock file or something. You can try to change the balancer and see if that makes any difference.

The only thing else I can think of is to give you a setting that ignores the lock file - but that could potentially have all kind of other consequences.

Hi Nicolai,

We don't have a balancer on this index (SecondayUsers) but I have noticed that the Products one (that does have a balancer) shows the same symptoms from time to time.

I initially suspected that the time interval between runs of the scheduled tasks is not long enough but I have noticed that the SecondayUsers indexing process takes less than a minute and the interval between runs is 1 hour.

I will keep it under observation for now and if nothing works, I will get back to you for that setting :)

Thank you,

Adrian

I am pretty sure this is related to how the disks are managed in their setup.

On Azure web apps the same issue can exist because of the disks are optimized for reads and not writes giving some issues on high load writes like indexing is.

BR Nicolai