Issue #9301: Users running into timeouts with error messages that are unhelpful - Pulp

Actions

Send by e-mail Copy link

Issue #9301

closed

Users running into timeouts with error messages that are unhelpful

Added by dalley about 3 years ago. Updated almost 3 years ago.

Status:

CLOSED - CURRENTRELEASE

Priority:

High

Assignee:

ipanova@redhat.com

Category:

Sprint/Milestone:

3.17.0

Start date:

Due date:

Estimated time:

Severity:

2. Medium

Version:

Platform Release:

OS:

Triaged:

Yes

Groomed:

Sprint Candidate:

Tags:

Sprint:

Sprint 106

Quarter:

Description

See attached bugzilla.

There was some additional discussion with a separate upstream user, who was running into this same issue: https://riot.im/app/#/room/#pulp:matrix.org/$xyV5D40cKUjqOYK7n7fk3J8v1RAqsu1KfafcYl968v8

The user had hypothesized that that maybe the "retries" were all counting towards the same "total timeout", but we have no confirmation of this.

This upstream aiohttp issue includes the same traceback and might provide some insight: https://github.com/aio-libs/aiohttp/issues/5239#issuecomment-727544671

I did also notice that we use the same ClientSession for all of a downloader's requests, that might be relevant: https://github.com/pulp/pulpcore/blob/master/pulpcore/download/http.py#L161-L163

Related issues

Actions

Copy link

Updated by dalley about 3 years ago

Description updated (diff)

Actions

Copy link

Updated by dkliban@redhat.com about 3 years ago

Triaged changed from No to Yes
Sprint set to Sprint 104

Actions

Copy link

Updated by dalley about 3 years ago

Priority changed from Normal to High

Assigning high because of multiple reports + BZ

Actions

Copy link

Updated by lmjachky about 3 years ago

Status changed from NEW to ASSIGNED
Assignee set to lmjachky

Actions

Copy link

Updated by rchan about 3 years ago

Sprint changed from Sprint 104 to Sprint 105

Actions

Copy link

Updated by lmjachky about 3 years ago

Status changed from ASSIGNED to NEW

Actions

Copy link

Updated by lmjachky about 3 years ago

Assignee deleted (~~lmjachky~~)

Actions

Copy link

Updated by lmjachky about 3 years ago

Just to note some findings:

The URLs which can be used for testing are: http://yum.oracle.com/repo/OracleLinux/OL8/appstream/x86_64/ and https://cdn.redhat.com/content/eus/rhel/server/7/7.7/x86_64/os.
You can also use python bindings to create a new remote object with certificates like this:

rpm_remote = rpm_remote_api.create({
    'name': RPM_REPO_NAME,
    'url': "http://yum.oracle.com/repo/OracleLinux/OL8/appstream/x86_64/",
    'policy': 'immediate',
    'client_cert': open("/home/vagrant/devel/13_sept_certs/rhcdn.crt", mode='r').read(),
    'client_key': open("/home/vagrant/devel/13_sept_certs/rhcdn.key", mode='r').read(),
    'tls_validation': False,
})

on_demand syncing does not trigger any erroneous behaviour.
immediate syncing ended up with "No space left on device" errors on my machine (not sure whether this error was caused by insufficient RAM space or regular disk space).

Actions

Copy link