Backup fails/incomplete

[JosephParticipant]

Purchased migration addon and tried it right away. Cannot get Backup to complete.

I’ve searched through the forum and find no help. Following is the problem I’m having.

I go to Settings>UpdraftPlus Backup/Restore> and select “Backup Now”

I monitor the “last log message” area. This log file seems to repeat itself over and over but never completes. Examples of what displayes is as follows:

Last log message = Adding batch to zip file (UpdraftPlus_BinZip): over 500 files added on this batch (11 Mb, 2035 files batched, 501 (502) added so far); re-opening (prior size: 62.1 Kb) (Mar 05 17:18:50)
Backup in progress = “Wed, March 5, 2014 17:18 – next resumption: 1 (after 45s) – show log – delete schedule”

Then
Last log message = Adding batch to zip file (UpdraftPlus_BinZip): over 500 files added on this batch (3.9 Mb, 2035 files batched, 501 (501) added so far); re-opening (prior size: 6916.3 Kb) (Mar 05 17:23:57)
Backup in progress = Wed, March 5, 2014 17:18 – next resumption: 2 (after 334s) – show log – delete schedule

Then
Last log message = Adding batch to zip file (UpdraftPlus_BinZip): over 500 files added on this batch (6.9 Mb, 2035 files batched, 501 (501) added so far); re-opening (prior size: 7931.7 Kb) (Mar 05 17:31:07)
Backup I progress = Wed, March 5, 2014 17:18 – next resumption: 3 (after 139s)

This repeats over and over–as many as 9 times.

Then “Backup, logs & restoring” may show a “set” available, but the backup in incomplete.

What, if anything, am I doing wrong?

[udadminKeymaster]

Hi Joseph,

I have some ideas, but I need to see the whole log file to be able to say something specific – please can you send it to [email protected]?

Best wishes,
David

[udadminKeymaster]

Hi Joseph,

The log file shows that the webserver is a bit slow (or perhaps just overloaded at the time of that backup – it’s not possible to distinguish these two possibilities from just the log file) configured to only allow PHP scripts (and hence WordPress/UpdraftPlus) to run for a maximum of 10 seconds. This is a toxic combination; there aren’t enough resources to get your website into a zip file with these limits. You should ask your web hosting company about getting the time limit raised – a normal allowance is 90 seconds or more,

Best wishes,
David

[JosephParticipant]

David,

I’ve run the plugin (successfully) on several other web site I do. I notice one difference right up front in the log. The log for all backups appears to ask “Requesting semaphore lock (fd)”. For the colonialshagclub.com backup log the next entries are:
Requesting semaphore lock (fd)
0000.016 (0) Semaphore (fd) was stuck, set lock time to 2014-03-07 19:28:42
0000.018 (0) Semaphore (fd) reset to 1
0000.019 (0) Set semaphore last lock (fd) time to 2014-03-07 19:28:42
0000.020 (0) Semaphore lock (fd) complete

For the other web sites the next entries are:
Requesting semaphore lock (fd)
0000.161 (0) Set semaphore last lock (fd) time to 2014-03-07 03:42:03
0000.162 (0) Semaphore lock (fd) complete

There appears to be some difference the host server’s posture.

I did contact the server’s tech support and they say I have 30 seconds to run. And it couldn’t be updated on the server I’m on.

[udadminKeymaster]

Hi Joseph,

Those log entries are what are expected if the previous backup attempt failed.

One other thing you can do is go into the ‘expert’ settings, and turn the ‘split’ size down to 25Mb. This will cause UD to split zip files more often, and help in low-resource situations. They’re saying there’s 30 seconds, but if you read the log, then you’ll see nothing is ever logged after more than 10.6 seconds in 10 attempts. So if there is 30 seconds available, then this means that it takes more than 20 seconds to do some quite simple zip file manipulations – that’s not an impossible scenario, but it would mean that the servers are either very slow by themselves, or very slow because of being overloaded.

David

[JosephParticipant]

Thanks — I ran it again (but not after implementing your above advice) and it took 25 “start agains” but it built the 4 zip files and a database. I moved those to a development location, used them to do a restore and it actually worked.

[Author]

Posts