# Stuck on "Starting …"

**URL:** https://forum.duplicati.com/t/stuck-on-starting/429
**Category:** Support
**Created:** [September 4, 2017, 10:21am UTC](https://forum.duplicati.com/t/stuck-on-starting/429 "2017-09-04T10:21:53Z")
**Posts on this page:** 1
**Showing post:** 13

<div class="post-metadata">

### Author: ![kenkendk](https://forum.duplicati.com/user_avatar/forum.duplicati.com/kenkendk/32/5305_2.png) [@kenkendk](https://forum.duplicati.com/u/kenkendk)
#### Post date: [September 15, 2017, 9:16am UTC](https://forum.duplicati.com/t/stuck-on-starting/429/13 "2017-09-15T09:16:32Z")

</div>

> [@NoLooseEnds](#):
>
> Why does the webui ignore the dbpath option?

Internally, the database path is stored in one place, and options in another. As many operations from the WebUI needs access to the database, it made sense to just hide the database things from the users.

I have been planning to fix this a bit by adding some logic that checks if the user has changed the `--dbpath` option and ask them how to handle it, but for now it is a bit obscure.

> [@NoLooseEnds](#):
>
> Does the temporary SQL file get very large? Or is it safe to leave it there if my /tmp has 4GB to spare?

That really depends on how much data you put in there. It should not grow larger than the size of the database itself (well, perhaps marginally if new data is added).

> [@NoLooseEnds](#):
>
> Could you please elaborate a bit on --check-filetime-only?

Duplicati compares: metadata, filesize, timestamp. If either of these values have changed, the file blocks will be scanned. The rationale is that hashing is an expensive operation, and can be avoided.

To get the previous values for this, it looks in the database and finds the previous values for the current file. Some users report that this lookup is slow. The `--check-filetime-only` makes a faster lookup, by only loading the previous timestamp, and ignoring the other values.

Some people are using programs that fiddle with the timestamps (i.e. TrueCrypt) and these files stay the same size and the same timestamp. You can use `--disable-filetime-check` to force scanning all the files regardless of their attributes.

> [@NoLooseEnds](#):
>
> And the current log file is full of:
> 
> ```plaintext
> Sep 14, 2017 11:30 PM: Starting - ExecuteScalarInt64: SELECT "ID" FROM "Block" WHERE "Hash" = ? AND "Size" = ?
> Sep 14, 2017 11:30 PM: ExecuteScalarInt64: SELECT "ID" FROM "Block" WHERE "Hash" = ? AND "Size" = ? took 00:00:00.015
> 
> ```

That is block lookups, it checks if it already knows the data that it found.

---

_[View the full topic](https://forum.duplicati.com/t/stuck-on-starting/429)._