rainloop cannot open mail folders with large quantities of messages

[fsantiago07044]

RainLoop version, browser, OS:

1.12.0, Chrome, Windows 10 Pro

Expected behavior and actual behavior:

Expected: i would click on my folder containing 26,243 messages, often with 3k+ unread messages or more, and it would open it and show me my messages.

Actual: rainloop produces an error for that folder stating: "can't get message list"

Steps to reproduce the problem:

1.> have mail folder containing large # of messages in total (unread count does not matter I feel because i can open the folder fine in roundcube mail and "read" all unread messages and rainloop still fails to list the folder contents).

2.> try to access said folder.

Logs or screenshots:

[ftoledo]

i hava same issue with courier imap server

[bschelst]

I have 200.000 unread messages in my inbox, and 1million+ emails in my inbox,, and rainloop works fine.

[fsantiago07044]

ok, what's your imap server?

-- Fabian A. Santiago

OpenPGP: 0x643082042dc83e6d94b86c405e3daa18a1c22d8f

Sent from my mobile.

---- On Sat, 20 Apr 2019 05:33:59 -0400 bschelst notifications@github.com wrote ----

I have 200.000 unread messages in my inbox, and 1million+ emails in my inbox,, and rainloop works fine. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

[marcecj]

I was also seeing this with my Nextcloud instance on my home server, using Dovecot for IMAP. Since I mostly use KMail and only use Rainloop from my laptop when I'm traveling, it took me a while to research the issue properly, but I finally did.

In my case the problem turned out to be that the default vsz_limit was too low for mmap() to be able to open the large mail folder's dovecot.index.cache file. Because of that, Rainloop was waiting for dovecot to scan through all of the folder's emails (>100000), causing it to time out with the error message shown in the screenshot above (if I waited for Dovecot's CPU usage to go back down I could usually reload the folder successfully). After increasing default_vsz_limit the error disappeared, and Rainloop performs much better now.

For comparison, here's some information on my setup:

• Rainloop 6.1.0 (run from inside Nextcloud 18.0.1, based on Rainloop 1.13.0)
• Dovecot 2.3.7.2 using the maildir mailbox format
• Gentoo Linux with a 5.4.20 kernel
• The server is a (used) Fujitsu Primergy TX140 S1 with 16 GiB of RAM, the hard drive on which the mailbox resides is, however, even older than the server.
• BTRFS with LZO compression and the autodefrag mount option (no qgroups).