Optimizing Oxidized config differ

Help
1

Edit: Disregard. This is entirely a Firefox thing. Chrome and Edge render instantly. Firefox seems to handle up to 10000 or so lines of config without too much indigestion, but go larger and the lag time is excessive.

First and foremost, I love the UX for the new config differ. Some nice work there. For the most part it works really well. I have a few exceptions using it to pull diffs on some rather large routers with ~20,000 lines or so of config – and thousands of config deltas in git. It takes a full minute for the differ to pull config plus deltas in and render the first diff.

Any ideas on how I could optimize for better performance? Thanks!

```

Component Version
LibreNMS 26.9.1 (2026-09-22T10:33:27-07:00)
DB Schema 2026_09_16_120000_widen_ent_physical_rev_columns (403)
PHP 8.5.4
Python 3.14.4
Database MariaDB 11.8.6-MariaDB-5ubuntu0.1 from Ubuntu
RRDTool 1.9.0
SNMP 5.9.4.pre2
===========================================

[OK] Composer Version: 2.10.3
[OK] Dependencies up-to-date.
[OK] Database Connected
[OK] Database Schema is current
[OK] SQL Server meets minimum requirements
[OK] lower_case_table_names is enabled
[OK] MySQL engine is optimal
[OK] Database and column collations are correct
[OK] Database schema correct
[OK] MySQL and PHP time match
[OK] Active pollers found
[OK] Dispatcher Service is enabled
[OK] Locks are functional
[OK] Python wrapper cron entry is not present
[OK] Redis is unavailable
[OK] No errors found with polling frequencies.
[OK] rrdtool version ok
[OK] Connected to rrdcached
```

2

So we call these two URLs at least when rendering:

/node/version?node_full=HOSTNAME_OR_IP&format=json

/node/fetch/GROUP/HOSTNAME_OR_IP

To your oxidized install, how long do those take if you run them in your browser or use curl?

3

Thanks for responding!

Ran node_full path json through jq to get number of lines.

$ time curl 'http://x.x.x.x/node/version?node_full=99k1&format=json' | jq > ox-node-full.txt
  % Total    % Received % Xferd  Average Speed  Time    Time    Time   Current
                                 Dload  Upload  Total   Spent   Left   Speed
100 942.1k 100 942.1k   0      0  5.64M      0                              0

real    0m0.198s
user    0m0.036s
sys     0m0.018s

$ cat ox-node-full.txt | wc -l
29724

$ time curl -s 'http://x.x.x.x/node/fetch/ASR-9K/99k1' | wc -l
20018

real    0m0.016s
user    0m0.004s
sys     0m0.009s

So retrieval time from Oxidized is almost instant via curl.

4

What about

/node/version?node_full=HOSTNAME_OR_IP&format=json

5 
$ time curl 'http://x.x.x.x/node/version?node_full=99k1&format=json' | jq | grep date | wc -l
  % Total    % Received % Xferd  Average Speed  Time    Time    Time   Current
                                 Dload  Upload  Total   Spent   Left   Speed
100 942.1k 100 942.1k   0      0  6.05M      0                              0
5404

real    0m0.184s
user    0m0.041s
sys     0m0.017s

jq with ‘grep date’ show 5404 commits for that node – and curl pulls the data in 2/10 of a second.

Edit: I’m thinking of revamping my git repos and putting them on a diet. I really don’t need four years worth of config deltas.

Edit 2: This is a Firefox issue. Diff is fast as can be using Chrome or Edge. SpiderMonkey vs V8 engine efficiency perhaps?

6

This only affects Firefox. Use Chrome or Edge if you get annoyed with lag on large config sets.

7

Thanks for doing the debugging. Hard to replicate this at the scale you have for me :frowning:

8

Quid pro quo. I was kicking myself a bit for not remembering Firefox’s javascript engine inefficiencies earlier. And yeah, some of our boxes are a little porky.