GNOME Bugzilla – Bug 754935
Error while Filtering new messages in 'INBOX'.
Last modified: 2015-09-17 06:17:30 UTC
When checking for mail, I am getting the following error message multiple times a day. Error while Filtering new messages in 'INBOX'. Execution of filter 'Junk check' failed: Empty cache file: /home/kip/.cache/evolution/mail/1441580502.4173.2@kip-desktop/folders/INBOX/cur/13/136733 This is strange because the file in the cache is not empty, but 16K. I have a "Junk" message filter that is there to catch what SA misses. If one of its rules is breaking, and there are many of them, there is no way to tell from this error message which rule it is.
Thanks for a bug report. I suppose this is an IMAPx account. It can be that the message download failed for some reason, and the referenced file was really empty in the time of the filter run, then another request made the file properly loaded. There used to be an issue with this for the IMAPx accounts. Since a fix for bug #745545 the internal IMAPx request processing had been made a bit more robust, also to address this particular error with message download. Please see also bug #737468, which is the same as yours. I'm marking this as a duplicate fo the later, but please note of the 'temporary' and 'workaround' words in its summary, the real fix is in the former bug report. *** This bug has been marked as a duplicate of bug 737468 ***
Thanks Milan. I will try to reduce the number of simultaneous connection threads to one to see if it makes any difference. Do you know which version of Evolution onwards this might have been patched? I am using 3.16.5.
The linked bugs at the comment #1 have the version numbers since which the changes are available, though those are development versions. The target in the stable releases is 3.18.0, to be released the next week.
Thanks Milan. I look forward to 3.18.0. In the mean time, however, I've noticed that the cache errors have disappeared since reducing the number of concurrent connection threads to one.
(In reply to Kip from comment #4) > the cache errors have disappeared since reducing the number of > concurrent connection threads to one. Right, that was one of the workarounds. I forgot to mention it here, I'm sorry.
No problem. Thanks again for your help. I'll let you know if 3.18.0 shows the same problem with N > 1 threads.